Comments Locked

395 Comments

Back to Article

  • shabby - Wednesday, October 27, 2021 - link

    Typo in the 12600 charts... 20 threads or 16?
  • nandnandnand - Wednesday, October 27, 2021 - link

    16 threads.
    2p + e = 2*6 + 4

    Thread count typos are going to happen a lot with Alder Lake, Raptor Lake...
  • at_clucks - Friday, October 29, 2021 - link

    20-30 years ago it was really easy to know everything important about a CPU (frequency more or less, maybe multiplier/FSB) just by looking at the model name. Intel made it really confusing for the past few generations even for the relatively mundane setup with multi-identical-cores at a certain frequency. Now with the different types of cores, different base and turbo frequencies, different thread count, etc. you'll have to search for the table every time.

    And we're in the easy part now, just high end, all models ending in "00". Can't wait for the "12672KSF" which has random tweaks here and there depending on what they could get out of the wafer.
  • yeeeeman - Friday, October 29, 2021 - link

    Neah, if you went to school at least you can count.
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    AMDead… until Zen 4. Zen 3+ is dead to me, because Zen 3+ is a band-aid solution to hobble along until they can get Zen 4 is out the door. Packing on a cache block is not going to beat this. Only a fresh and fast microarchitectural update can regain performance for AMD. And that update is Zen 4.
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    Also, timing is crucial and hitting the moving target. Zen 4 may beat Alder Lake but can it beat Raptor Lake which will likely come out late next year? If not, AMDead will be dead even with Zen 4, even if Zen 4 is a huge improvement over Zen 3. AMD has to compete against Intel’s best at time of release, not against itself.
  • schujj07 - Wednesday, October 27, 2021 - link

    Yes because we have independent benchmarks out already from real world usage and not Geekbench. Don't forget to look at the power usage on the new Intel CPUs. They very well could be faster than AMD, however, if they need double the power to be 15% faster that really isn't a win.
  • shabby - Wednesday, October 27, 2021 - link

    Don't think intel cares about losing the efficiency crown on the desktop, they just want to be at the top of the charts at all costs. Now in the server department with dozens of cores then it starts to bite them in the ass.
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    Most desktop users could care less about power efficiency. If this were a laptop or notebook question, that would be a totally different story. That is an interesting point you bring up, though! It's likely why Intel released desktop BEFORE mobile. Historically, they release first to mobile and then desktop. Power efficiency may be very well why they opted to avoid pushing it to mobile first. Or not at all. They may just be trying to counter AMD where it hurts most, in their current Ryzen stronghold.
  • schujj07 - Wednesday, October 27, 2021 - link

    Power efficiency does make a slight difference on the desktop. If you need use a huge AIO or air cooler compared to a mid range air cooler that hurts your pocket book. That can also make your system louder than the comparable other system. Not to mention it will affect your summertime electric bill if it keeps heating up the room so your AC is kicking on.
  • DougMcC - Wednesday, October 27, 2021 - link

    You don't have a wall port so your desktop can vent heat directly outside?
  • ballsystemlord - Wednesday, October 27, 2021 - link

    Ha ha ha! Good one!
    We all need one of those now that the high end draws 100s of watts.
  • tamalero - Thursday, October 28, 2021 - link

    I always laugh at these type of comments. Because when intel eats a ton of energy, its "indifferent". because "desktop".
    But when AMD chunks a lot of energy.. its all about how bad AMD is XD
  • Qasar - Thursday, October 28, 2021 - link

    yep, intel does it, it fine, if amd does it, people lose their minds and castrate amd for it, go figure
  • kwohlt - Wednesday, October 27, 2021 - link

    "Or not at all"

    Between Laptop, Desktop, and Server...Desktop is the least important market. There's no way Intel wouldn't release Alder Lake on Mobile
  • boozed - Wednesday, October 27, 2021 - link

    "Most desktop users could care less"

    How much less?
  • Qasar - Wednesday, October 27, 2021 - link

    " Most desktop users could care less about power efficiency. "
    you sure about that ? i have a space heater beside me for the winter if needed, it says " Noma " on the side. i dont want one that says " intel inside "
    intel needs to throw power efficiency out the window in order to compete with Zen, if they didnt, they would probably still be losing across the board
  • Flying Aardvark - Thursday, October 28, 2021 - link

    He's right. If you're worried about power efficiency, buy a laptop or a tablet. There's docking stations. Why even bother with a desktop at all if you're not going to go all out for max single threaded performance.
  • Spunjji - Thursday, October 28, 2021 - link

    "Why even bother with a desktop at all if you're not going to go all out for max single threaded performance."
    This is literally not a requirement for something that probably amounts to 98% of desktop users.
  • Icehawk - Thursday, October 28, 2021 - link

    I’d agree the a typical consumer doesn’t care about this but I do - using AMD allows me to use a smaller PSU, small enough it’s fanless. I go for quiet and powerful on my home machines.
  • Zoolook - Saturday, October 30, 2021 - link

    On the contrary, I have built a number of systems for other people and maximum power is never requested, however, everyone wants a quiet desktop, even for gaming it's all about the GPU unless you "compete" with extreme fps shooters.
  • hansmuff - Thursday, October 28, 2021 - link

    Why bother? Because I want a quiet machine, and laptops are not quiet under heavy load. And I want a fast GPU that runs reasonably quiet (undervolted.) Laptops and great and all, but they're not the clear alternative to quiet desktop computing.
  • Flying Aardvark - Thursday, October 28, 2021 - link

    hansmuff- anything can be downclocked. If you're going to gimp your system, then why get the one with less potential if you change your mind? Intel has more potential in gaming if you go all out. Otherwise get your tablet.
  • Spunjji - Friday, October 29, 2021 - link

    Flying Aardvark - all indications so far are that AMD retains better perf/watt. We'll have to wait to see whether or not that applies to gaming, but it's disingenuous to suggest that one need "gimp your system" just to get acceptable thermals. It never used to be necessary with Intel, either, before they started clocking the nuts off their processors to keep up with Ryzen.

    "Otherwise get your tablet."
    Doesn't merit a response
  • Flying Aardvark - Saturday, October 30, 2021 - link

    Spunjji- but Intel is faster, all out. It makes no sense to neuter any system, you should be buying what has the most potential. Again, if you want the best perf/watt, get your M1 or your tablet of choice. Best 'perf per watt'. I'll be on an i9, which is faster than your system. And I can afford the extra $1 a month electric bill for the <5% time it's on and consuming high power.
  • TheinsanegamerN - Friday, October 29, 2021 - link

    Because a laptop with a dock is not only more expensive but also compromises cooling and noise compared to a desktop, doesnt ahve the expandability of a desktop and.....honestly why am I even trying to explain this. You knwo full well there are use cases for desktops beyond "ZOMG 1% IPC INCREASE GOOO 1000W COOLER"
  • TheinsanegamerN - Friday, October 29, 2021 - link

    Incorrect. Less efficiency means more heat/noise, more expensive coolers, and the need for more expensive motherboards. As a desktop owner Id otn want to waste money trying to cool down a hot running SOB.
  • PedroCBC - Saturday, October 30, 2021 - link

    I found it really funny that americans think that electric bills are the same all around the world, as if they were even within their own country. Yeah buddy, you are right, everyone can put a 1600W PSU and run it full throttle all day and pay without it take a reasonable amount of money.
  • Flying Aardvark - Saturday, October 30, 2021 - link

    PedroCBC- but you wouldn't be running at full throttle all day. Think!
  • yeeeeman - Friday, October 29, 2021 - link

    I don't think they will be double the power...for example 12600K needs 150W top TDP and it beats the 5800X which is at 142W...
  • Qasar - Friday, October 29, 2021 - link

    yea, according to leaks maybe. remember how the 1100 series went ?

    untill reviews are out, its all speculation
  • Josh128 - Wednesday, October 27, 2021 - link

    Intel's only chance to beat Zen 4 is if they get their "Intel 4" 7nm process going. If they have to stay on "Intel 7" 10nm, thats not going to cut it against TSMC 5 /5 +.
  • regsEx - Wednesday, October 27, 2021 - link

    Why calling Intel 4 as 7 nm process and Intel 7 as 10 nm process, but not mentioning TSMC N5 as 9 nm process?
  • Spunjji - Thursday, October 28, 2021 - link

    Two reasons:
    N5 was never renamed
    "Intel 7" is still not equivalent to TSMC N7 on either density or power characteristics. Who knows how "Intel 4" will compare to N5 - if it's a similar relationship to Intel 7 and TSMC N7, then it should be called Intel 6.
  • shady28 - Thursday, October 28, 2021 - link

    Intel 7 aka 10ESF is about the same density as TSMC N7FF+. It is reportedly 100.76 MT/mm2 vs TSMC N7FF of 96.5 MT/mm2. That's easy to validate doing your own research. This doesn't mean that Intel is not behind TSMC 5nm, but it is no longer behind their 7nm.

    https://semiwiki.com/semiconductor-manufacturers/s...

    https://fuse.wikichip.org/news/1371/a-look-at-inte...
  • Spunjji - Friday, October 29, 2021 - link

    @shady28 - those are the official figures, sure. It's a little more difficult to validate those, but luckily I have done my own research!

    In reality neither Intel nor TSMC can hit their claimed densities in actual chips, for a variety of reasons. The best information we have suggests that AMD have managed 62.8MTr/mm² on TSMC N7 with Renoir (Cezanne is an oddball because it has blank die space). Meanwhile Intel's competing chip - Ice Lake - clocked in at 53.2MTr/mm² on 10nm+. It's not a vast difference, but it's a difference all the same.

    What we don't know is how the 10 SF and 10 ESF (now Intel 7) process changes have affected density. There's circumstantial evidence to suggest that 10 SF involved a relaxation of density, but we don't really know because Intel have been extremely tight-lipped about it.
  • Zoolook - Saturday, October 30, 2021 - link

    In efficiency they sure are i.e performance/watt on the CPU's produced they are way behind.
  • melgross - Friday, October 29, 2021 - link

    It’s really more like a 7nm Intel process.
  • Samus - Friday, October 29, 2021 - link

    What's important of note here is AMD has had a fantastic run, effectively beating Intel in most categories for the last few years. Now Intel may be caught up and they are neck and neck.

    But this wouldn't have happened unless Intel had a fire under their ass.
  • melgross - Friday, October 29, 2021 - link

    The way AMD had a fire under their ass
  • DigitalFreak - Wednesday, October 27, 2021 - link

    Fanboi says what?
  • Kangal - Friday, October 29, 2021 - link

    What?

    But to get a little serious, I don't think Intel is going to win with their big.LITTLE architecture. I feel like ARM has a huge lead on the 15W (or less) demographic. So it would make sense for x86 designers to double-down on their performance lead in the higher thermal envelope. That's what AMD is (seemingly) going for with its focus on lower-latency Infinity Fabric, +5nm node to clock higher, and their 3D-Stacking of Cache. Not to mention all the help from DDR5, Pcie 5, nVme, Wifi 6 etc etc.

    Intel's approach will win them back the Laptop segment, but they won't be winning the tablet segment back from ARM. And even the Gaming Laptop segment won't be an outright victory against AMD's offerings, not to mention the New MacBook Pros. If anything, Intel should have capitalised on their Atom efficiency cores, and do little.BIG computing in like 2018.

    Servers is a position where Intel may see improvements. But it's still in favour of AMD for now and the near future. The bigger threat comes from next-gen ARM-servers. I doubt anything from the left-field will come, RISC-V is still a paperlaunch/niche for the next few years.

    So while I think Intel is (FINALLY) becoming competitive against AMD, I don't think they have enough to go on. Their node is still inferior. Their Xe-Graphics are still inferior to RDNA-2. And they still lag behind AMD's Cores when you factor in Infinity-Fabric and 3D-Cache. Not to mention that the system/kernel is not quiet optimised yet (let alone individual programs) when thinking about Windows11.

    For now, we have to choose from:
    Android, iOS, macOS, Windows
    RISC-V, ARM, Apple-ARM, Intel, AMD.
    ARM-Mali, PowerVR, Apple-Graphics, Nvidia, AMD RDNA.
  • Silver5urfer - Thursday, October 28, 2021 - link

    What is this fanboy junk...sigh.

    ADL demands Windows 11 POS, you want to shill for the HW which demands installing a strictly mobile junk copied OS with zero respect to computing factor on top where they are saying VBS is mandatory on all OEM machines and purposefully nuked AMD L3 performance ?

    I have a positive opinion on this ADL but it has insane changes, like Intel ITD drama who wants to endure that band aid solution of Intel with 2 layer system in between the OS and CPU. On top the major issue being socket longevity. How long this socket will retain it's value and will Intel release another Z790 next year ? No idea.

    Now for your AMD bashing, Zen 3 wiped the floor with 2 generations of processors yeah they have bugs while OCing and DRAM tuning, but if you run at stock no issues and performs very well competitively. And for the ADL performance, it's honestly a joke. Because ADL has small trash cores since Intel wants to sell more BGA junk and they cannot beat the performance with more cores due to 10nm heat.

    Raptor Joke lol so ADL CPU is going to be EOLed under a year lmao, just like 11900K ? 2 CPUs in succession. While 10900K still stands strong. That's Intel for you. Meanwhile AMD's Zen 3 is now ready for 2022 action as well with AM4 and 3D V Cache. Keep using the yearly socket refresh and chipset refresh and CPU refresh while coming here and spout nonsensical load.

    Finally pay up DDR5 tax and premium premature trash DDR5 quality, by 2023-2024 DDR5 will be matured and all ADL buyers will weep hard.

    Now for the closure. Zen 4 is going to steamroll over Raptor Joke, 100% garunteed. Do you think these companies operate without knowing what their competitor is doing ? they operate 2 years ahead of cycle internally. Plus AM4 experience is very important for AMD to fix the bugs from Platform to CPU. Ultimately they cleared out saying we are not using joker big little design. A full far Zen 4, massive price increase is also coming from them, the IPC boost and the ST SMT performance will send fanboys to darkages.
  • Silver5urfer - Thursday, October 28, 2021 - link

    I forgot to post important thing, be happy that you have AMD as competition else Intel would have been selling you 4C CPUs even in 2021 and AMD is pushing x86 to next level, if that design dies or stagnates PC will die. Keep the x86 alive if you want to own a computer not a consumable garbage ARM product.
  • MaxIT - Thursday, October 28, 2021 - link

    That works both ways: AMD dominion is not welcomed in the same way. Did you see what AMD did with prices ? AMD and Intel are the same: when they think they are above competitors, they start taxing customers. Let them fight to prevail: we customers will be the winners
  • Qasar - Thursday, October 28, 2021 - link

    " Did you see what AMD did with prices ? " you referring to the $50 price increase between zen 2 and 3 ? 50 bucks is nothing, compared to how much intel kept raising their prices over the years before zen 1 was released. but yet, very few complained about that.
  • Oxford Guy - Friday, October 29, 2021 - link

    That $50 is a response to the inflation that has been happening from all of the Covid money printing.
  • mode_13h - Friday, October 29, 2021 - link

    It's not only money-printing. There are legit shortages due to outbreaks in factories, and worker-shortages in certain sectors.

    I suspect one reason for the trucker shortage, in the US, is that truck drivers tend to be older and overweight, which are both risk factors for complications from Covid-19 (which the nature of their job also increased their exposure towards). So, I truly wonder how much the US truck driver shortage is due to drivers unable to continue performing their duties due to complications (or death).
  • Spunjji - Friday, October 29, 2021 - link

    @mode_13h - It's a good point. A lot of chatter about the effects of COVID seems to ignore how many people more than usual died. It's not world-war levels of death, but systems subject to stress have to eat into margins to cope, and a lot of the world's financial and supply-chain systems were already under stress from tariffs and sustained economic strife when COVID hit - so there weren't a lot of margins left.
  • mode_13h - Saturday, October 30, 2021 - link

    @Spunjji a lot more people have long-term effects from Covid-19 than the ones who died. Death is just the worst outcome, but there are many people unable to function at the same level as before. And I'm not only talking about "long Covid", where the immune system seems to be stuck in an overstimulated state, but other sorts of cardiovascular and organ damage it can cause.
  • Oxford Guy - Friday, October 29, 2021 - link

    Things are rarely ‘only’. It’s also very rare for anyone to mention the inflation from the money printing.
  • mode_13h - Saturday, October 30, 2021 - link

    Just to nit-pick, a lot of it isn't money-printing. What qualifies as money-printing is the bond buying programs by US Treasury and possibly other central banks. However, a lot of what's going on is debt-financed government spending.
  • melgross - Friday, October 29, 2021 - link

    AMD’s prices were less because few wanted their chips. As that began to change, so did their prices.

    There’s nothing nefarious about AMD or Intel pricing. It’s simple supply and demand.
  • Qasar - Friday, October 29, 2021 - link

    melgross, here you are crying about amds prices, were you also crying when intels prices kept going up ? some how i doubt that very much.
  • melgross - Wednesday, November 3, 2021 - link

    I’m not crying about either. Where do you get the crying part from? What I said is true. Don’t get so emotional about this.
  • Spunjji - Friday, October 29, 2021 - link

    Until Tiger Lake H, Intel *were* still selling 4C dies as "premium" products in 2021 😅
  • Sivar - Wednesday, October 27, 2021 - link

    Real-world benchmarks are hell on hypotheses. Wait for the reviews.
  • OreoCookie - Wednesday, October 27, 2021 - link

    That’s very short-sighted: AMD and Intel have staggered releases so that one can claim the performance crown for a while for a few months until the other releases new products.

    Besides, Zen3+ seems to be a very good bandaid. If AMD’s performance claims are to be believed (and they have been very accurate with their Zen-line of processors), Zen3+ will be competitive with Alderlake but have much lower power consumption. So Intel will still be far behind in terms of efficiency and be roughly on par in terms of performance.

    Efficiency matters greatly because Intel is using the same core design across machines. Lower efficiency means that they can’t scale up their server CPUs to the same degree AMD can. And that laptop CPUs are still less efficient, i. e. lower battery life at equivalent performance.
  • whatthe123 - Wednesday, October 27, 2021 - link

    the only thing AMD claimed is about 15% general uplift in the games they tested. they are pretty reliable, but they also didn't make any claims about anything else, so it doesn't really make sense to expect any more than that unless announce it.
  • haukionkannel - Thursday, October 28, 2021 - link

    Cache will affect only certain applications. But games actually could benefit from it (not all games, but some).
    So all depends on what you do with your computer and what does it cost... I am expecting Zen3+ to cost much more than alder lake! Much much more... But Alder lake motherboard are expected to cost quite a lot and ddr5 memory whyle not as expensive as I did expect are still 50% more than ddr4... So in total Zen3+ may be sensible upgrade to some Zen owners!
    For total ddr5 build I would wait Zen4 and maybe even Intel upgrade after Alder lake to see if we get decent ddr5 by then and almost normal price. Maybe only 15% extra over ddr4 in that time period. First gen new is.... well expensive for those who has money to jump to something new as soon as it is available. Second and third gen are where we have more mature tech and hype is less affecting the prising!
  • MaxIT - Thursday, October 28, 2021 - link

    You are right , but Lisa Su now needs a reality check. Her price policy was outrageous, because she knew she had the best technology available (Zen 3) last year.
    Now 5600X and 5800X prices are ridiculous.
    AMD needs to return at least at Zen 2 levels, in term of prices.
  • Qasar - Thursday, October 28, 2021 - link

    you do understand that a lot of the prices we are seeing are because of the shortages, crypto miners, higher demand because of covid etc. again, amd only increased its prices by $50, MSRP, the rest is do to the other factors.
  • Oxford Guy - Friday, October 29, 2021 - link

    Inflation is high.

    American wildflower seed (the real thing, not florists' stuff) has doubled in price for many species. Some have gone higher than that (tripled over several years' time). Covid hasn't caused a shortage but inflation is causing the prices to go way up. This was happening even prior to Covid.

    Wildflower seed is unlikely to be the only sector that has seen massive price jumps.
  • mode_13h - Friday, October 29, 2021 - link

    All seed prices surged, because the pandemic fueled a renewed interest in gardening. Also, flowers and houseplants make for good video conferencing backgrounds, which probably contributed to the demand surge (and therefore price inflation) of those products.
  • Oxford Guy - Friday, October 29, 2021 - link

    mode, you’re fond of stating claims about things you don’t know about as if they’re fact rather than speculation.

    The dramatic price increases in wildflower seed were happening prior to Covid. Covid money printing exacerbates the problem.

    This seed is not used for gardening primarily. Next time either do enough research or stick to topics you have knowledge in.
  • mode_13h - Saturday, October 30, 2021 - link

    What I know is that seed prices surged due to pandemic-fueled interest in gardening. I also know that houseplant demand has surged, as well. I have no specific knowledge of wildflower seeds, but when they're adjacent to two other areas that have experienced a demand surge, it's a reasonable assumption to make that similar forces acted on that market.

    Also, while inflation is notable, it's not nearly enough to account for what you're seeing. You accuse me of stating claims about things I don't know, but that's exactly what you did by attributing the price increases to simple inflation.
  • Oxford Guy - Sunday, October 31, 2021 - link

    'I have no specific knowledge of wildflower seeds'

    Yes.
  • Qasar - Friday, October 29, 2021 - link

    if you really think inflation is the cause of prices going up, then good for you. i doubt inflation went up that much to cause the price increases we have seen aka video cards
  • Spunjji - Friday, October 29, 2021 - link

    @Qasar - it didn't, but some people seem to favour oddly specific explanations over more generally plausible ones.
  • Oxford Guy - Friday, October 29, 2021 - link

    Inflation has been happening prior to Covid but a lot of money was printed in response to it.
  • melgross - Friday, October 29, 2021 - link

    You’ve got it backwards, prices going up result in inflation. That includes chip prices.
  • Qasar - Friday, October 29, 2021 - link

    not as much as they have, this time there are other factors
  • Oxford Guy - Friday, October 29, 2021 - link

    It’s fallacious to use this ‘sole factor’ argument. Of course there are multiple factors. Talking about all of the important ones is wise, rather than the select topics that are constantly focused upon.
  • TheinsanegamerN - Friday, October 29, 2021 - link

    6 core went from $159 to $299 for a 20% performance boost. Unforgivable.
  • MaxIT - Thursday, October 28, 2021 - link

    This. I hope Zen 3+ will show some others architecture improvements, 3D cache aside, to compete with Alder Lake. If we are speaking just about more cache, several applications won’t gain anything.
  • Spunjji - Friday, October 29, 2021 - link

    Zen 3+ on desktop reportedly got cancelled (if it ever existed). It looks like we're going to get it on mobile, but desktop is going to be Zen 3 and VCache in some as-yet unknown configurations.
  • Spunjji - Thursday, October 28, 2021 - link

    Calling it a "band-aid" is a bullshit marketrolling technique anyway. It's a technology AMD developed in concert with TSMC on their own timescale - the fact that it may enable Zen 3 to better compete with Alder Lake is icing on the cake, not the intent of the design.
  • yetanotherhuman - Thursday, October 28, 2021 - link

    Sweet, where have you found benchmarks, or are you just talking out of your arse?
  • Spunjji - Thursday, October 28, 2021 - link

    How do you know this?

    Oh. You don't.
  • sing_electric - Thursday, October 28, 2021 - link

    I mean possibly but, at the same time, it looks like the Zen 3+ will be a drop-in replacement that could fit in an existing build, so the total spend is just the processor, meaning users could get a decent performance boost over say Zen 2 for like $300.

    That's pretty alluring compared to going to Alder Lake (especially with DDR5, where you'd likely be spending $700+ all told).
  • dotjaz - Friday, October 29, 2021 - link

    And you are irrelevant.
  • Chaser - Friday, October 29, 2021 - link

    I think the news of the decade was AMD overtaking INTEL. Now it's just "faster" pissing contests. The winner here is the consumer.
  • DannyH246 - Wednesday, October 27, 2021 - link

    Hahahaha worst CPU ‘launch’ since Intel’s last one. More hot as hell crap. Can’t wait for proper benchmarks to show how crap it really is.
  • Bobbyjones - Wednesday, October 27, 2021 - link

    Sweet baby jesus this looks like an amazing launch. Intel seems like it will retake the performance crown in gaming. They look like they will catchup to AMD in productivity which is a HUGE improvement. And the prices undercut AMD?!? I am on Zen 1 right now but screw it, im preordering a 12700k. Intel won me back.
  • shabby - Wednesday, October 27, 2021 - link

    They won you back based on this preview? Maybe wait for the review before preordering...
  • Makaveli - Wednesday, October 27, 2021 - link

    lol I was thinking the same thing.
  • Spunjji - Thursday, October 28, 2021 - link

    There's a lot of, ahem, new commenters who've suddenly popped up on various sites to tell everyone about their pre-order for a product that has no reliable performance information out. I still remember what Intel said about Rocket Lake before launch. This should be better, but I'm not giving them the benefit of the doubt.
  • tamalero - Thursday, October 28, 2021 - link

    Not only forums, also youtube. The names almost seem like randomly generated.
    Sounds like intel's PR team doing the IMDB and RottenTomatoes way (aka buying reviewers).
  • Qasar - Thursday, October 28, 2021 - link

    in some ways, i hope the performance claims of 12th gen, dont pan out, just to see what all the intel fanboys say and the excuses they come up with :-)
  • AhsanX - Friday, October 29, 2021 - link

    You will be the one inhaling Copium
  • Spunjji - Friday, October 29, 2021 - link

    We already saw how that went with Rocket Lake - they instantly shut up and/or start talking about the Next Big Thing.

    12th gen looks like it'll be a solid improvement on what went before it, I'm just so damn tired of Intel apparently having so little faith in their engineering achievements that they feel the need to hire spammers (who in turn make the genuine fanbots feel validated).
  • Qasar - Friday, October 29, 2021 - link

    Spunjji the same was said about 11th gen. and it kindof fell flat.
    untill the reviews are out, its just smoke and bs from intel pr, like always
  • Bobbyjones - Wednesday, October 27, 2021 - link

    Unfortunately the days are over of waiting for reviews. If you didnt preorder or werent in line for Zen 3, RDNA 2, or Ampere, you had to wait MONTHS to get what you wanted. Im not making that mistake again.
  • shabby - Wednesday, October 27, 2021 - link

    When was the last time an intel cpu was out of stock on launch day?
  • Drumsticks - Wednesday, October 27, 2021 - link

    I mean, probably the last time they had a good release.
  • shabby - Wednesday, October 27, 2021 - link

    You mean the 2500k? :)
  • tuxRoller - Wednesday, October 27, 2021 - link

    If you have needed something faster that badly why haven't you already bought something to replace that three year old cpu?
  • Eagle07 - Wednesday, October 27, 2021 - link

    Bobby.... maybe read the fine print on the intel numbers. They are penalizing AMD by 15-20% on gaming performance. They used the windows 11 build that introduced tripled AMDs L3 latency and ignored prime cores. They also gave it super slow ram with high cas. Wait for reviews unless you want to be had by Intel's marketing team.
  • DigitalFreak - Wednesday, October 27, 2021 - link

    Nah, let Booby have his fantasy.
  • drothgery - Thursday, October 28, 2021 - link

    Real world impact on games is more like 5% according to Tom's Hardware.
  • boozed - Wednesday, October 27, 2021 - link

    Preordering CPUs now, what a time to be alive.
  • Gigaplex - Wednesday, October 27, 2021 - link

    I bought my Zen 3 from retail on the day of launch without preordering.
  • Spunjji - Thursday, October 28, 2021 - link

    "Unfortunately the days are over of waiting for reviews"
    Only for a fool

    "Im not making that mistake again."
    Just before you said you're still on Zen 1... derp.
  • shady28 - Thursday, October 28, 2021 - link

    Already happening with Alder Lake too, pre-orders opened up yesterday afternoon at places like Best Buy and mostly sold out in minutes. 12900K / KF and 12600KF pre-orders are sold out on Newegg, leaving the 12700K 12600K and 12700KF.
  • Qasar - Thursday, October 28, 2021 - link

    being sold out is meaningless, until it is confirmed how many were available TO pre order. for all any one knows there could of been only a few hundred available. if that

    just checked one store here, and you can still pre order the 12th gen cpus.
  • Spunjji - Friday, October 29, 2021 - link

    Who cares. Wait a few more weeks, there'll be more stock. If it's not at a price you like, keep waiting. Very few people have requirements that can't be met by 5+ year old hardware - even gamers. If they do, then they should already have bought into something like Zen 3 (plenty of stock for months now) or Comet Lake on discount.
  • Qasar - Friday, October 29, 2021 - link

    heh, i just picked up a 5900x for 667 cdn this week
  • schujj07 - Wednesday, October 27, 2021 - link

    Until there are independent benchmarks we do not know how well this is going to perform. Before Rocket Lake came out there were leaks of synthetic benchmarks that had it destroying Zen 3. Well the real world benchmarks came out and guess what Rocket Lake was only marginally better than Comet Lake. One thing to note is that Intel is once again having to put power consumption the way of the Dodo. Most likely so that the chip is able to compete with AMD's vastly more efficient uArch.
  • Gothmoth - Wednesday, October 27, 2021 - link

    and the energy you save with such a great 241W CPU..... ;-)
  • Gothmoth - Wednesday, October 27, 2021 - link

    oh sorry 12700k is your choice.... that´s just 190W.
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    Most people don't care about your $10.00 in energy savings.
  • Teckk - Wednesday, October 27, 2021 - link

    Why even go with big little/power and efficiency cores in desktop then if energy doesn’t even matter?
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    I agree there. I would much rather see all performance cores on a desktop SKU. big.LITTLE makes more sense in a mobile application where battery life and standby power are important.
  • The Hardcard - Wednesday, October 27, 2021 - link

    But isn’t about more power? Aren’t nearly all workloads that can peg more than six cores parallelized? What exceptions are there?

    So you have 4 E- cores in the space of one P - core and those E - cores boost muti-threaded performance. When iwould a ninth or higher P -core count do better?
  • nandnandnand - Wednesday, October 27, 2021 - link

    The efficiency cores are essentially Comet Lake cores but at lower clocks and no hyperthreading. They can fit many of them onto the die. For highly multi-threaded workloads, the more efficiency cores the better. Which is why Intel will be doubling them for Raptor Lake, and maybe doubling them again after that.

    Intel is betting you won't need more than 8 performance cores because they will handle up to 8 (or 16) single-thread sensitive tasks. Even games that can use at least 8 cores probably don't treat all of the threads as if they need the highest clocks 100% of the time.

    It could take multiple generations before Intel has a good reason to bump performance cores to 12.
  • Spunjji - Thursday, October 28, 2021 - link

    This has been mentioned enough times that people should know, but to be clear, the E cores are there to boost multithreaded performance without bloating the die size.
  • kwohlt - Wednesday, October 27, 2021 - link

    Because given the choice between 10 pCores OR 8 pCores + 8 eCores (given the same power and thermal constraints), the 8+8 build is not only offering better performance in most workloads, it's also more scalable (see Raptor Lake doubling eCores again next week, so 13600K = 6+8, 13700K = 8+8, 13900K = 8+16).
  • Mikewind Dale - Wednesday, October 27, 2021 - link

    I could see big.LITTLE being beneficial for corporations with large numbers of computers, where everyone is using basic applications like Chrome, Word, etc. The efficiency cores could save the corporation a lot of electricity and cooling costs.

    But for workstations performing tasks like video-editing, the efficiency cores may not be worthwhile.
  • nandnandnand - Wednesday, October 27, 2021 - link

    I could see businesses using the cheaper 6-core die (i5-12400) with no efficiency cores.

    The Alder Lake 2+8 ultra mobile die put in an all-in-one or small form factor mini PC would also be a good choice. That die has up to 96 graphics EUs.

    Workstation users should eventually be using efficiency cores. Just imagine if Intel made a CPU with 8-16 performance cores and 128-256 efficiency cores. If an application can scale to use hundreds of cores, you want hundreds of efficiency cores.
  • Teckk - Thursday, October 28, 2021 - link

    The peak power consumption doesn't seem to fit the case. Why even go for this for basic apps, where an i5 would do just fine?
  • Flying Aardvark - Thursday, October 28, 2021 - link

    That's a design decision that's focused on mobile. Clearly it won't hurt gaming performance. BTW AMD is moving to hetero cores too, they're simply behind the curve right now.
  • Gigaplex - Wednesday, October 27, 2021 - link

    I care about the noise levels.
  • Spunjji - Thursday, October 28, 2021 - link

    I'll happily also save $100 on the CPU cooler, though... and that's before we get to the extra platform costs. Those are some pricey boards thanks to the early adopter tax and chonky VRMs needed, and DDR5 is paaaiiinfully expensive.
  • factual - Thursday, October 28, 2021 - link

    Whether the energy savings is $10, or less, or more, or none at all is unknown. What Anandtech should do is measure the power consumption of Alder Lake while playing various games and compare it against power consumption of Zen3 while playing the same games. That would be a realistic power consumption comparison. Most people rarely fully load their CPUs, so comparing power consumption while playing demanding games is the most useful metric IMO.
  • Spunjji - Friday, October 29, 2021 - link

    If you're a gamer, for sure. If you're not, those numbers aren't terribly useful. The problem is that it increases their workload substantially and doesn't necessarily give a clearer picture.
  • Meteor2 - Monday, November 1, 2021 - link

    Lots of people here are gamers. And rather obviously, just extend the test to compiling, rendering etc
  • Josh128 - Wednesday, October 27, 2021 - link

    Lol, you can likely upgrade to a Zen 3 + using your same mobo in a couple months, needing only a CPU. Going Intel will require $1000 for new mobo, RAM, and CPU. Think about that for a second.
  • Tilmitt - Wednesday, October 27, 2021 - link

    He probably can't upgrade because AMD are deliberately sabotaging older motherboard support for their newer CPUs.
  • Gigaplex - Thursday, October 28, 2021 - link

    No they're not.
  • kobblestown - Thursday, October 28, 2021 - link

    Well, they tried to. Sure, at the end they listened to customers rather than to MoBo makers (who are also their customers, BTW) but this cannot be taken for granted. Look what they did to the ThreadRipper platform.
    Don't get me wrong, I'm typing this on an 3600X/X470, my laptop has 5700U and my server is TR1920X. My next CPU will also be AMD by the looks of it. But let's not pretend they are "nice". They are a company and will do everything to maximize their profits. Being the underdog, they have to play it nice. But once their marketshare started growing they took some pages out of Intel's book. The only thing that can keep them in check is competition. That's why I root for Intel in this case. Of course, Intel are effectively lying about AlderLake's performance (using unpatched Windows 11 - how conveniently that crippled Zen's performance!) but the faster ADL turns out to be, the better for AMD customers.
  • Spunjji - Thursday, October 28, 2021 - link

    "Well, they tried to"
    So they're not, then. When's the last time Intel listened to customers and enabled support for newer chips on an older board?
    ...
    ...
    ...
  • cfenton - Thursday, October 28, 2021 - link

    Sabotaging is too strong, but they did drop support for first gen Ryzen boards. I had a 1700x and couldn't go to a 5800x without a new board, so I got a 3800xt at a pretty reasonable discount. Three generations is better than two, but it's still not great.
  • Spunjji - Friday, October 29, 2021 - link

    On the flip side, Zen to Zen 2 is a greater performance (and perf/watt) increase on a single socket than Intel had been offering across multiple socket changes for years before Zen showed up.

    It's disappointing, but it's also so much better than what we've had to put up with for a while now.
  • MaxIT - Friday, October 29, 2021 - link

    When was last time they did it ? Never
  • Flying Aardvark - Thursday, October 28, 2021 - link

    Intel never lost the gaming crown other than older titles like CSGo. Early reviews were borked and didn't have ABT enabled on the 11900K. That's still the king. The 11600K can nearly match the i9 if overclocked, its been the best value going.
  • Qasar - Thursday, October 28, 2021 - link

    you sure? various reviews say other wise.
  • Flying Aardvark - Thursday, October 28, 2021 - link

    Qasar, do you know what you're looking at when you read those? Most had ABT disabled, and disabled MCE (a default bios option that no one changes, especially gamers). Take a look again.

    https://www.tomshardware.com/reviews/cpu-hierarchy...

    https://www.eurogamer.net/articles/digitalfoundry-...
  • Qasar - Thursday, October 28, 2021 - link

    going AT bench, in games the 11700k loses more often then not in games to the 5800X as seen here : https://www.anandtech.com/bench/product/2775?vs=26...
    and it doesnt look much better for the 11900k either :
    https://www.anandtech.com/bench/product/2787?vs=26...

    the toms link is the geomean results, or average results across the whole test, individually, intel lost the gaming crown with zen 3.
    anandtechs bench, seems to say other wise
  • Flying Aardvark - Saturday, October 30, 2021 - link

    I haven't vetted AT's benches, but Tom's are properly configured with ABT enabled and MCE at default. If you win on average, you have the faster gaming chip.... that's how averages work. Sorry the results aren't what you've been brainwashed to believe is reality. i9 still holds the crown and in fact, never lost it to AMD.
  • Qasar - Sunday, October 31, 2021 - link

    sorry, but your the brain washed one here, its known fact that amd has the gaming crown over all with zen 3 and 11th gen. intel isnt wining on average, sure intel may win a game here and there, but over all, it loses, look at the reviews them selves, not the aggregate rankings like you are referencing.
    " The Core i9-11900K delivers impressive performance in both single-threaded applications and gaming. Still, it doesn't cement itself as the clear leader in gaming and also trails the price-comparable (at least based on general pricing guidelines) Ryzen 9 5900X in threaded workloads. "
    from: https://www.tomshardware.com/reviews/intel-core-i9...

    11900k review : https://www.anandtech.com/show/16495/intel-rocket-...
    for the most part the 11900k lost in most of the gaming tests, in some cases even it even lost to the 5600k !!!! and you claim intel never lost the gaming crown ?? look at the INDIVIDUAL reviews, not the geomean average results. per game intel loses, this is fact. but keep believing that you want, " I haven't vetted AT's benches " well them maybe should should give them a look. but keep believing what you want, even though you are wrong.
  • Flying Aardvark - Monday, November 1, 2021 - link

    Your AT review didn't have ABT enabled. THG is valid and it trades blows and wins in the aggregate. What else do you want? The average consists of per-game results. You're wrong.
  • Spunjji - Thursday, October 28, 2021 - link

    Behold! More shiny new Intel FUD accounts
  • Flying Aardvark - Thursday, October 28, 2021 - link

    Spunjji your ignorance must be bliss. Happy to run my 11900K against your potato in Cyberpunk or any AAA game of your choice anytime. Keep it moving chump.
  • Spunjji - Friday, October 29, 2021 - link

    Bold words there, but zero content given you had no idea who I am or what hardware I own. As it is, I run a laptop based on a 4800H, so I have no doubt it would lose to an 11900K - but if you fancy a challenge, try limiting that chip to 45W and then see how well it competes.

    I paid £700 for my entire system - 4800H, RX 5600M, 16Gb RAM - and it does everything I need. By consequence I don't feel the need to throw insults at other people in a vain attempt to justify my ill-considered purchase.
  • Flying Aardvark - Saturday, October 30, 2021 - link

    I don't buy PCs to downclock them, but you would probably still lose as RKL IPC is the best in the business until Alder Lake lands. No worries for me, I'll have a 12900K as well. And a 13900K etc. I'm not buying based on price, I buy the fastest gaming CPU available.
  • Meteor2 - Monday, November 1, 2021 - link

    Zen 3 has higher IPC than RKL.
  • Flying Aardvark - Monday, November 1, 2021 - link

    Meteor2- there's many ways to measure IPC but in general, it doesn't. RKL has the highest IPC on the market today.
  • Qasar - Monday, November 1, 2021 - link

    prove it
  • Flying Aardvark - Monday, November 1, 2021 - link

    Prove what? That there's many ways to measure IPC? Are you stupid or just ignorant? You have to measure IPC per-task. If you want an example run Geekbench on your little engine that could and share your single threaded score.
  • Qasar - Monday, November 1, 2021 - link

    and you are wrong, rkl has has WORSE IPC then zen 3 :
    https://hitechglitz.com/leaked-benchmarks-show-tha...

    https://www.tomshardware.com/features/intel-core-i...
    and i quote " The outcome of our Ryzen 9 5900X vs Core i9-11900K battle seems pretty straightforward on the surface: The Ryzen 9 5900X wins four out of five categories, while also scoring a tie in the features section. Our CPU faceoff ends up being a five to two win in favor of the Ryzen 9 5900X, meaning that the choice should be quite clear for most enthusiasts.

    The Ryzen 9 5900X ultimately wins on the strength of its better blend of gaming and application performance, not to mention that it comes with much lower power consumption that ultimately results in a cooler and quieter system. And that's despite it coming with four more cores than the 11900K. "

    https://arstechnica.com/gadgets/2021/03/intel-11th...

    https://www.guru3d.com/news-story/leaked-benchmark...

    enough said, you are the stupid and ignorant one. nice try, but the results are there IF you care to look at them, but i bet you wont. you will still reference the same geomean results as if it was gospal.
    rkl has WORSE IPC then zen 3m a simple google search for " which has higher IPC Zen 3 or rocket lake " would of shown you that. granted a some of them are leaked benchmarks, but there are others, like toms that i posted above, that show zen 3 has the better IPC.
  • MaxIT - Thursday, October 28, 2021 - link

    At least wait for a review
  • SarahKerrigan - Wednesday, October 27, 2021 - link

    That 12600K does not look half bad. Looking forward to reviews.
  • maroon1 - Wednesday, October 27, 2021 - link

    All alder lake parts are good.
  • willis936 - Wednesday, October 27, 2021 - link

    Check's in the mail.
  • shabby - Wednesday, October 27, 2021 - link

    More like a free intel inside sticker 😂
  • Makaveli - Wednesday, October 27, 2021 - link

    lmao yup the intel fan boys are out by the truck load today.
  • Spunjji - Thursday, October 28, 2021 - link

    Indeed it does not! If it performs like the leaks indicate it should be the best i5 since the 2500K and finally give AMD a reason to drop prices on the 5600X. I'm cautious, though. I can't remember the last time Intel undercut on price *and* had a performance advantage.
  • maroon1 - Wednesday, October 27, 2021 - link

    Intel will destroy AMD this time

    Core i5 12600K (12600KF), not only destroy 5600X easily but it also beats 5800X
  • nandnandnand - Wednesday, October 27, 2021 - link

    If this slide is accurate, AMD can come back for a tie or smaller loss with Zen3D:

    https://cdn.videocardz.com/1/2021/10/Intel-Core-i9...

    Zen3D will work on existing AM4 motherboards.

    After that, it's Zen 4 destroying Raptor Lake.
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    Raptor Lake is also promising big gains over Alder Lake so you have to keep that important piece of information also in mind. Zen 4 doesn’t just have to beat Zen 3 and Alder Lake. It has to clear the bar that Zen 3, Zen 3+, Alder Lake, and Raptor Lake will have set with sufficiently enticing pricing and performance.
  • nandnandnand - Wednesday, October 27, 2021 - link

    Raptor Lake will only have slight gains on the Performance cores. It will double Efficiency cores to 16, but that won't matter for gaming. Zen 4 should crush it on single-threaded and gaming performance. For multi-threaded, depends if they give it a third 8-core chiplet or not.

    Pricing is what it is. They can lower prices or "jebait" as needed. Zen 3 has already been available under MSRP, and it could go down more. For Zen 3D, if AMD can get close to matching Alder Lake, they will try to keep prices as high as they can get away with since it will be the final upgrade for AM4 users.
  • maroon1 - Wednesday, October 27, 2021 - link

    nandnandnand,

    You aware that P cores in Raptor Lake will use new architecture ?? It is called Raptor Cove, and I expect it to be faster than Golden Cove.

    Also, Zen4 need around 20% boost to match 12900K single-thread....... To crush 12900K, it need a lot more than 20%.... ..And yet you claim that Zen4 should crush even Raptor lake..LOL... Good luck with that.
  • nandnandnand - Wednesday, October 27, 2021 - link

    Zen3D = +15% in games or applications where the tripled L3 cache is useful, which should bring it close to Alder Lake. Zen 4 should be about +25% IPC over Zen3/Zen3D, with some models also having 3D V-Cache.

    I only expect +10% going from Golden Cove -> Raptor Cove.
  • Makaveli - Wednesday, October 27, 2021 - link

    Wasn't the reported IPC uplift for Zen 4 to be 29% and that is the pre-vcache version.
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    Yes, from Chips & Cheese.
  • Spunjji - Thursday, October 28, 2021 - link

    "I expect it to be faster than Golden Cove."
    Fanboy expects newer chip to be faster. Who could have predicted this stunning revelation. Raptor Lake will still be on Intel 7 (10ESF / 10+++) so it's not getting any easy gains.
  • guycoder - Wednesday, October 27, 2021 - link

    Well you know what they say. Only real men have real cores...But seriously competition is good even if you are an Intel or AMD fan boy. Upshot is we are going to get more cores, faster speeds and cheaper prices no matter what team you support. Intel always had a great engineering team. It has been shackled by poor manufacturing and myopic marketing for a long time so it's good to see them come out with something new and not a refresh. I will still wait out this generation and being early adopter of what is going to be very pricey supporting hardware. Zen 4/4+ vs. Meteor Lake with AVX-512, ML/AI accelerators and low latency DDR5 when supply and demand are sorted is a smarter bet.
  • maroon1 - Wednesday, October 27, 2021 - link

    1- I don't trust sides/tests from either company (including AMD claims about Zen3d)

    2- I was not talking about gaming..... Gaming is not the only thing CPU does... When I said 12600K will be faster, I was talking about the general performance.
  • Spunjji - Friday, October 29, 2021 - link

    "I don't trust sides/tests from either company"
    I love me some "both sides" goodness. AMD slides previewing Zen, Zen+, Zen 2 and Zen 3 were all proven to be accurate by independent reviews. Intel's have been reliably inaccurate.
  • MaxIT - Friday, October 29, 2021 - link

    That’s only for gaming. A bigger cache could have less of an impact on other applications
  • Gothmoth - Wednesday, October 27, 2021 - link

    241W TDP in intel speak means 280W under full load.

    you pay twice the money per year for energy..... and how do you cool this thing?
  • Wrs - Wednesday, October 27, 2021 - link

    Most of us don't sustain our processors at peak. Average consumption for fixed work and lightly threaded/idle power are substantial inputs to energy cost, which is probably not the biggest consideration for a desktop. If it were you'd get a laptop?

    On cooling, if you're getting a 241W turbo processor you are not aiming for a low-profile build. Any $70 tower cooler ought to handle 241W with ease, if the processor interface is strong. Intel's have historically been strong. AMD is usually behind there. For example the 5800X package can only dissipate around 160W on the best liquid/air coolers, as the power density is too high on the CCD and the solder is thick.
  • Spunjji - Thursday, October 28, 2021 - link

    More like $100 to handle 250W "with ease".

    "Intel's have historically been strong"
    Not between Ivy Bridge through to (some) Comet Lake.

    "For example the 5800X package can only dissipate around 160W on the best liquid/air coolers"
    You don't *need* to dissipate more than that, though? You barely get more performance than running it around 120W.

    Really loving this game of talking up Intel's strategy for their CPUs producing absurd amounts of heat. Like great, they deal with heat better, *because they have to*. Inefficient is as inefficient does.
  • Wrs - Thursday, October 28, 2021 - link

    No, $70 for 250W+. There's a somewhat hard limit for heat pipes. A $100 cooler is typically dual tower or equivalent with a limit in the 350-450W range (though admit I never set out to measure those specifically). That doesn't mean there aren't crappy designs, but I'm referring to any reputable maker. That also doesn't mean a cooler will cool to the limit on any processor. There are choke points with thermal interfaces and die size. Back in '08 I had this single tower, two-fan Thermalright 120 extreme that successfully sustained 383W on LGA1366 at under 100C... that's an above average $100 cooler. Might have gone higher with a bigger/thinner die but just illustrates the possibilities.

    The 5800x, on the other hand, cannot practically sustain over ~160W on that caliber of cooler. In fact a high-end heat sink (I've used a D15 and U12S) remains cool to the touch while the CCD is throttling at 90C, with some 135W over 88mm2, simply because the thermal interface down below is the choke. On the double CCD Zen 3's, the amperage at the socket seems to be the limit. Plenty of people know Zen3 doesn't have as much overclocking headroom, simply by the difference between stock 1C turbo and all-core OC frequency, forcing users to pick between snappy ST performance and sustained MT. Notice how both Rocket Lake and ADL don't have that issue, given sufficient cooling.

    Lastly, let's not confuse efficiency with rated power limits. The review sites will have to measure ADL efficiency empirically. From a theoretical view I don't spot a big difference between Intel 7 and TSMC 7nm, so to see the Intel 7 part rated for so much higher power than the N7 part (all the best Zen3's are 142W turbo) tells me that Intel's package/process accommodates much higher heat dissipation and by extension has more room to perform better whether stock vs. stock or OC vs. OC. And it's kind of expected based on the physical characterization of a 208mm2 monolithic die (per der8auer) with a reduced z-height and thin solder, as compared to Zen 3's typically thick package and thin IHS.
  • Oxford Guy - Friday, October 29, 2021 - link

    '$70 for 250W+'

    Noisy.

    Let's look at the cost per watt for a quiet installation.
  • Wrs - Friday, October 29, 2021 - link

    Maximum noise for a cooler is based on the fans and airflow path through the cooler, not the heat. The duty cycle - and thus noise - for typical PWM fans is regulated based on processor temperature, again not actually the heat. So if you want quiet, get a quiet fan, or get a processor with good enough efficiency and thermal dissipation or heat tolerance that it won't need the fans at 100%.

    Hope I didn't overcomplicate the explanation. When I put a $70 stock Noctua U12S on my 5800x and start a game, it gets somewhat noisy. That's because the ~100W being put out by the CPU isn't dissipating well to the heatsink, not because 100W is a challenge for a U12S.
  • Spunjji - Friday, October 29, 2021 - link

    That's more a function of how you have your fan curves configured. If the CPU isn't putting out enough heat to saturate the heatpipes, and the die temp is going to be high no matter how fast you run the fans because of thermal density, then you have room to reduce the fan curve.
  • Oxford Guy - Friday, October 29, 2021 - link

    Coolers that are undersized (and less expensive) make more noise.

    It’s similar to the problem with Radeon VII. The die was designed to be small for profit and the clock had to be too high to try to compensate.

    Quiet cooling costs more money in high-wattage parts. It’s not complicated, beyond the fact that some expensive AIOs are noisier than others.
  • mode_13h - Sunday, October 31, 2021 - link

    > Radeon VII. The die was designed to be small for profit and
    > the clock had to be too high to try to compensate.

    Radeon VII was not designed as a gaming GPU. It was small because it was the first GPU made at 7 nm, by a long shot. At that time, it was probably one of the biggest dies made on that node.
    The fact they could turn a profit by selling it at a mere $700 was bonus.

    And the MI50/MI60 that were its primary target don't even have a fan. They have a rigid power limit, as well. So, the idea that AMD just said "hey, let's make this small, clock it high, and just run the fans fast" is a bit off the mark.

    https://www.amd.com/en/products/professional-graph...
  • Spunjji - Friday, October 29, 2021 - link

    Maybe we live in different places. Where I am, a decent tower capable of cooling 250W comfortable - not maxed out - is the equivalent of $100 US.

    "The 5800x, on the other hand, cannot practically sustain over ~160W on that caliber of cooler"
    I don't know why anyone would bother, though. The difference between MT performance with PBO and overclocked MT performance is minimal. If you need more MT that badly and TDP isn't a problem, then Threadripper is a better option. If your use case doesn't cover the cost of Threadripper then it's unlikely you'll miss a few percent in performance and you'll probably benefit from not overspending on cooling just to get it. Rocket Lake doesn't compete well with Zen 3 in MT even when overclocked, so it's not a great argument for that chip. We'll have to see how it pans out with ADL, though it does look promising.

    "Lastly, let's not confuse efficiency with rated power limits"
    I'm not!

    "tells me that Intel's package/process accommodates much higher heat dissipation"
    Sure, but...

    "by extension has more room to perform better"
    ...this is absolutely not something you can deduce just from knowing that it can sustain higher power levels 😅
  • Wrs - Saturday, October 30, 2021 - link

    @Spunjii It's a detail I bother with. If two chips are equally performant at a fixed wattage, but one chip can sustain a higher wattage, then it is fair to deduce that chip as having more performance potential. This is something we'll have to dissect between ADL and Zen 3 of course. Their process nodes suggest they ought to be similarly efficient at fixed power, but there will be other factors as OS, different cores, and quantity of cache.

    @Oxford Guy The goalposts are moving. Any reputable $70 cooler is not outrageously loud. If you need extra quiet then you'll pay up or tweak the fans regardless of whether you got Zen 3 or ADL or RKL. I was just pointing out that Zen 3 has its own dissipation problems (due to design & build) that could set it back in performance on the same heatsink.
  • Wrs - Saturday, October 30, 2021 - link

    Also, Noctua agrees with my findings. Here you can search a wide variety of processors they have tested with most of their coolers: https://ncc.noctua.at/cpus/model/Intel-Core-i9-129...

    According to Noctua, the vanilla U12S gets 3 bars for overclocking headroom on the 11700/11900k and 12700/12900k, but only 2 bars for Ryzen 3700x/5800x/5900x/5950x, despite all of us knowing how much more heat RKL dissipates compared to Zen 2/3. (Noctua has access to ADL samples that we don't.) The weaker thermal interface on the Ryzen MCMs is real; it wasn't just me misapplying thermal paste.
  • Oxford Guy - Sunday, October 31, 2021 - link

    'The goalposts are moving. Any reputable $70 cooler is not outrageously loud.'

    Apparently. Somehow 'noisy' is not 'outrageously loud'.
  • Wrs - Sunday, October 31, 2021 - link

    @Oxford If you have a problem with the design noise of $70 coolers or air cooling in general then say so lol. I'm just pointing out that at stock everything, a typical $70 tower cooler on a Ryzen 5800 or higher nets you maximum fan noise under full CPU load, even on an open bench, even as the heat pipes remain cool to the touch. That's a thermal conduction deficiency in the processor package.

    You can lower fan speeds, change out the fans, undervolt/throttle the processor. But that's true regardless of what CPU you got.
  • MaxIT - Friday, October 29, 2021 - link

    Well, to be honest they are not speaking about TDP anymore. 241W is the peak power consumption under load, according to their data.
  • DigitalFreak - Wednesday, October 27, 2021 - link

    Man, the Intel fanbois are getting major boners over leaked synthetic benchmarks. They're so desperate they'll take anything they can get. Kinda like incels...
  • Spunjji - Thursday, October 28, 2021 - link

    Also like Incels: they project, constantly. Everything is always the fault of "AMD fanboys".
  • Spunjji - Thursday, October 28, 2021 - link

    blah
  • psychobriggsy - Thursday, October 28, 2021 - link

    'destroy' is a bit strong.

    It looks like ADL will have the single core lead, and finally fixes Intel's multi-core problem. However it might not be by much, the power use looks like it will be higher, and there is the platform and DDR5 cost and early adopter issues.

    AMD's big worry is the 12600KF, which may also perform excellently on DDR4 due to losing some cores. It might mean that the 12600KF solution exceeds the 5800X at a similar cost to the 5600X. AMD have had a while to think of a solution to this (IMO price reductions and a 105W higher-clocked version of the 5600X might help), but they might just release all this with the v-cache products in two or three months rather than rushing a band-aid out now.
  • drothgery - Thursday, October 28, 2021 - link

    Intel has no intention of destroying AMD even if they could; keeping anti-trust lawsuits away is helpful. Keeping AMD in a position where they're forced to compete on value rather than performance, though, they're all for that.
  • trivik12 - Wednesday, October 27, 2021 - link

    Why is alder lake not compared with AMD APU and instead compared with CPU without IGP?
  • shabby - Wednesday, October 27, 2021 - link

    The apu's are neutered.
  • nandnandnand - Wednesday, October 27, 2021 - link

    Most Alder Lake gamers are going to use a discrete GPU, especially for these top SKUs. Even with the crazy GPU prices.

    If anyone uses the iGPU, it will be the type who buys the (unannounced) i5-12400 with 6+0 cores, or i3-12300/i3-12100 (4+0). The kind of chip that will land in cheaper prebuilt/office PCs.
  • drothgery - Wednesday, October 27, 2021 - link

    I suspect most will use the iGPUs, but they'll be using the mobile SKUs with 96 EUs, not the desktop SKUs with 32 ... and they won't be playing AAA games that way.
  • nandnandnand - Wednesday, October 27, 2021 - link

    "Thread Director: Windows 11 Does It Best"

    Nobody should use Windows 11. Switch to Linux.
  • prophet001 - Wednesday, October 27, 2021 - link

    big true
  • Wereweeb - Wednesday, October 27, 2021 - link

    Yep. I've had enough of Bloatsoft's spyware.
  • Spunjji - Thursday, October 28, 2021 - link

    Currently trialling it and it's a big meh from me.
  • Oxford Guy - Friday, October 29, 2021 - link

    'Nobody should use Windows 11. Switch to Linux.'

    Agreed. Windows 11 is a test of the passivity of consumers. If consumers reward MS for orphaning a lot of performant relevant hardware it will know it can be Apple Jr. That's not a good thing for your wallet.
  • Spunjji - Friday, October 29, 2021 - link

    It's really easy to install 11 on older hardware. That's not its biggest problem.
  • GeoffreyA - Friday, October 29, 2021 - link

    Indeed, it's been shown to work on a Pentium 4. So all it really needs is an SSE2/x64 CPU. That's it.
  • Oxford Guy - Friday, October 29, 2021 - link

    We’ll see. Things have a way of changing after one gets a foot in the door. Incrementalism is the standard tactic for pushing bad changes onto people.

    You may not take MS’ support policy seriously but many do. My recollection of the initial coverage of this issue is that people were warned sternly by the installer if not blocked.

    There is no excuse for trying to bully people into abandoning equipment that has good performance and relevance.

    Claims about what’s easy abound. For instance, many claimed that it’s easy to get macOS Big Sur to run on specific Dell hardware, via the ‘Dortania guide’ and packages on Github. The reality is a far different matter.

    The only truly easy thing is to have legitimate support from MS. Instead, the company is trying the same underhanded tactics it used to push Windows 10 onto machines. The repertoire now includes arbitrary removal of official support from highly-relevant machines.

    No excuses.
  • mode_13h - Friday, October 29, 2021 - link

    > Nobody should use Windows 11. Switch to Linux.

    In the short term, Alder Lake users will be worse of for it. Intel hasn't been prioritizing work on properly supporting the Thread Director & Alder Lake scheduling in Linux.

    I expect that will change, but their current priority is clearly Sapphire Rapids (for which Linux support has been running uncharacteristically late).
  • domih - Wednesday, October 27, 2021 - link

    Pop-corn is ready. Bring it on. Can't wait to see this discussion degenerate into a slugfest.
  • onewingedangel - Wednesday, October 27, 2021 - link

    The pcie 4.0 x4 is a bit of an odd choice - it's SSD's that are likely to benefit from pcie 5.0 first, so you'll end up needing to run the pcie 5.0 x16 as x8/x8 with 8 for the GPU and x8 for an SSD adapter. If you split the x16 can 8 lanes be pcie 4.0 (if using current GPU) with the other 8 running pcie 5.0 for SSD?
  • willis936 - Wednesday, October 27, 2021 - link

    The I/O situation looks like a disaster. But hey, they give the users 8 SATA ports.

    🙄
  • Wereweeb - Wednesday, October 27, 2021 - link

    PCIe 5.0 is worthless for consumers. The real world performance bottleneck for SSD's is random reads/writes.

    14GB/s sequential reads will only help when transferring things like a game installation to another PCIe 5.0 SSD - and as the new SSD's will need bleeding edge microcontrollers to not fry themselves, I can guarantee that very few people will consider it worth it to buy more than one. And to those people, I would strongly recommend getting their hands on an Optane drive instead.

    Will all of that change at some point? Naturally! But it will take years for applications of those sequential speeds to be found, and for prices to come down.
  • Spunjji - Thursday, October 28, 2021 - link

    Agreed. PCIe 4.0 was already of dubious value, 5.0 is just ridiculous. I guess it's "futureproof" though.
  • Oxford Guy - Friday, October 29, 2021 - link

    SSD makers want everyone to use QLC crud. They're not prioritizing performance.
  • PEJUman - Wednesday, October 27, 2021 - link

    "...In my mind, this is a bad oversight. I was told that this is explicitly Microsoft’s choice on how to do things."

    I thought Pat came out a few weeks ago admitting software is a key component to performance. It seems the rest of rank-and-file have not gotten the memo and still try to shift blame towards Microsoft.

    Coming from a user perspective, I really don't care who own the issue. If I paid $3000 for a high performance laptop, I really don't have the time to figure out how to extract maximum performance and stability out of my workstation. I don't want to have to google which W11 update I need to have, then google again on why Windows keep rolling back that update from my system.

    At this rate Apple ecosystem with guaranteed performance, reliability & better QA will eventually win my next laptop purchase from Dell/HP.
  • romrunning - Wednesday, October 27, 2021 - link

    I think the universal recommendation will be to use the "High Performance" power plan on all desktops. Then you don't have to worry about the threads being shifted onto E-cores if you really needed it on a P-core.
  • PEJUman - Wednesday, October 27, 2021 - link

    I agree this is easy, but that's not the point.

    What I am saying is how much longer will you tolerate this kind of quality? why should I fiddle with power profiles to patch a broken/nonQA'd scheduler. Microsoft does not pay me for beta testing their scheduler, they also failed to pay me to beta testing their thunderbolt 3 and 4 implementations. And to make this worse, This is a product that MS actually sells and tries to make money from.

    I do not have to do any of these with on the macbook Air. And the macOS is freaking free, it's licensed to their hardware set.

    FYI here is what I currently running, just in case you're wondering:
    Desktop: 5950x + 3090 @ 8K on HDMI 2.1 Homebrew
    Laptops: Dell Inspiron 1165g7, XPS 1065g7, HP Zbook workstation with i7 6th gen. All of these crashes repeatedly with TB3 and TB4 docks from Dell/HP/OWC. And guess what, the fix is not to let the laptop sleep (sounds familiar?).

    Apple: phones, ipads, macbook air with 8th gen i5 + TB3 dock.

    These apple products have much higher uptime, almost 20x better than the MS products above. My desktop is by far the most stable, but still a long shot away from the mac. Looking at this article, I expect W11 with Adler lake laptops to go even worse. Intel, AMD and MS need much tighter integration and QA to compete with M1s from Apple. Microsoft, Intel and AMD, if you are reading this. Next year, I am betting that my money will be spent towards a M2 & MacOS powered laptop. Please prove me wrong.
  • Robberbaron12 - Wednesday, October 27, 2021 - link

    The Thunderbolt 3 and 4 Implementation on Windows 10 has been one giant Charlie Foxtrot. We have had endless issues with Dell and HP laptops and desktops with terrible TB software and drivers crashing continuously. M$ and the OEMs blame each other and nothing improves (I'm actually pretty sure its Intels firmware) but Apple can make it work so ????
  • Spunjji - Thursday, October 28, 2021 - link

    Intel and MS are the two consistent factors on the PC side. Could be MS, could just be Intel writing lousy drivers. Hard to say.
  • Dug - Wednesday, October 27, 2021 - link

    The Macbook Air M1 release was a clusterf. Memory management was hosed creating GB's of writes a day to ssd. TB docks did not work and caused kernel panic. External monitor support was non existent, meaning you couldn't control resolution or refresh rate on most popular monitors. I know because I lived through the beta testing and release. So don't go thinking Mac is all grandiose all the time.
  • PEJUman - Wednesday, October 27, 2021 - link

    Is this still a problem today? Will it be fixed by the time M1max/pro hits meaningful quantities in the wild?
  • Spunjji - Thursday, October 28, 2021 - link

    Fixed, AFAIK
  • Oxford Guy - Thursday, October 28, 2021 - link

    The shattering screen hasn't been fixed.
  • name99 - Thursday, October 28, 2021 - link

    "Memory management was hosed creating GB's of writes a day to ssd."

    And yet the only people who ever cared about this were people who insisted it meant early death of the SSDs and were looking for something to be wrong with the machines.
    I am unaware of a single case where this had any real-world effect, and as far as we know, it may well have been bugs in the SW that was reporting these numbers.

    "I know because I lived through the beta testing and release."
    What do you expect from beta testing?
    If you'd stuck to "I know because I lived through the initial release", and dropped the stupid "OMG my SSD will be dead in three months" hysteria, you'd be a lot more convincing. As it is, you come across as the sort of person who insists on finding things to complain about, and if you can't find something reasonable, you'll find something unreasonable.
  • Oxford Guy - Thursday, October 28, 2021 - link

    Apple has yet to fix the CD player bug I reported back in 2001.

    The original Mac OS (last released as OS 9) played audio discs at 1x. OS X has always spun the discs at the maximum read speed of the drive, which is utter incompetence.

    I just checked and Catalina is still doing it.

    I reported the bug via Apple's OS X report page at least twice, probably four times — over the years. That a $1 trillion company can't manage to get audio CDs to play at the correct speed in decades is beyond appalling.
  • Oxford Guy - Friday, October 29, 2021 - link

    Uptime is definitely an area where macOS/OS X excels and Windows does very poorly.
  • schujj07 - Wednesday, October 27, 2021 - link

    Is it just me or does Z690 only catch up to X570 that has been available for like 2 years?
  • Spunjji - Thursday, October 28, 2021 - link

    It has PCIe 5.0, for whatever that's worth, but yeah in other regards it's just catching up
  • Gothmoth - Wednesday, October 27, 2021 - link

    pffffhhhhhhhh.............. 241WATT ROTFL
  • Gothmoth - Wednesday, October 27, 2021 - link

    really nobody commentimng on the TDP.. especially of teh 12900k?

    when intels puts 241W in the whitepaper you can be sure that´s already downplayed.

    people who buy such a CPU need the processing power and often use it under full load for hours.
    this thing should be boycotted by greta fans :)

    honetsly i am curious to see the thermals....
  • Wrs - Wednesday, October 27, 2021 - link

    Never mind the CPU's TDP, every gamer with a mid-range discrete or better worries about a boycott from Greta fans. 241W is timid by Rocket Lake standards! Plus I'd think these boards let you set custom limits.
  • Oxford Guy - Thursday, October 28, 2021 - link

    If you think Thunberg is the problem facing humanity you need a reality check.

    Or, perhaps have a chat with the record number of extinct species we can pat ourselves on the back for.
  • Spunjji - Friday, October 29, 2021 - link

    I tend to switch off as soon as someone invokes Greta Thunberg in a disparaging way. There are plenty of reasonable discussions to be had about the limits of that kind of activism, but I've yet to see anybody who specifically invokes her name come close to having one. There seems to be a real desire to take a massive and worrying problem and slap it all onto one scapegoat.
  • Wrs - Saturday, October 30, 2021 - link

    @Oxford Missing the point. They sell you a CPU just like they sell you a car. You get a certain design efficiency and peak performance. It's up to you how you wanna use that. Full throttle it a lot and you'll get a high energy bill.

    Talking about extinctions check out Ordovician. Stuff happens on Earth. It can certainly be our problem
  • Oxford Guy - Sunday, October 31, 2021 - link

    'Stuff happens on Earth.'

    That's the point I missed?
  • Wrs - Tuesday, November 2, 2021 - link

    The point isn't about Thunberg. Whatever she says does not change the laws of nature. If she gets us to think in a more educated way about the impact of all our little decisions, that’s a good thing. If she gets us to discover more about the physical processes happening around us, that’s awesome. Just know that identifiably man-made problems are not necessarily the ones we need to solve. We fight forest fires caused by lightning. Vaccinate animals against viruses. If need be, I believe we’d try to fragment an asteroid heading our way. How does our impact on Earth compare to that 10-mile-wide rock 66 million years ago? Still standing by humanity’s record penchant for causing extinctions?

    A CPU is a tool to do some mechanical thinking. How we fabricate, power, and use it can leave traces on Earth. So is the case for everything else we have. Those traces can accumulate and come back to bite us. But it is not as simple as lower peak wattage is better. We need to do much more nuanced thinking than that.
  • mode_13h - Friday, November 5, 2021 - link

    > We fight forest fires caused by lightning.

    Focusing on what sparks the fires is missing the point. Forests are getting dried out by changes in rainfall and increasing heat. That's what makes the difference between the sort of containable forest fires we've seen in the past and the kind of out-of-control conflagrations we're seeing today.

    > Vaccinate animals against viruses.

    Wildlife? Really? You're going to vaccinate like a billion bats? And against viruses we don't even know about, yet? Good luck with that.

    > How does our impact on Earth compare to that 10-mile-wide rock 66 million years ago?

    We're living through the greatest mass-extinction since that time. Do you think that's okay?

    > it is not as simple as lower peak wattage is better.

    Some disagree with that. I'm less a fan of regulating the efficiency of energy sinks, however. I'd rather see traditional externalities priced into energy costs. Then, let the market do its thing and optimize energy use for us. Unfortunately, carbon pricing has never really caught on, in a big way.
  • shabby - Wednesday, October 27, 2021 - link

    Hey Ian will you be doing a ddr4 vs ddr5 comparison too in the review? Ddr5 prices seem to be double compared to ddr4 on Newegg now.
  • Roy2002 - Wednesday, October 27, 2021 - link

    Can motherboard makers put DDR4 slots along with DDR5 slots?
  • shabby - Wednesday, October 27, 2021 - link

    Seems like no, in the ddr5 section it says that intel requests oems not to do that.
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    OEMs, ignore Intel's stupid requests. We will buy DDR4 and DDR5 mixed slot motherboards!
  • Spunjji - Thursday, October 28, 2021 - link

    Bit late to be making that demand, tbh
  • OFelix - Wednesday, October 27, 2021 - link

    Looking forward to the day that Intel get their power draw under control:
    i9-12900K 241W
    R9 5900X 142W
  • Gothmoth - Wednesday, October 27, 2021 - link

    i am all for competition.

    but instead of praising the new intel architecture.... reviewers should have a close look if that´s not just a BRUTE FORCE approach to be competitive.
  • shabby - Wednesday, October 27, 2021 - link

    It seems since zen came out that's all intel has been doing, auto overclocking the cpu till the cpu hits 90c or so and keep it there. Can't blame them since there seems to be headroom unlike in the zen architecture.
  • OFelix - Wednesday, October 27, 2021 - link

    but only compared to when running weird AVX512 stuff that I won't use, I think?
  • OFelix - Wednesday, October 27, 2021 - link

    woops, meant to reply to the comment below :-)
  • Gigaplex - Thursday, October 28, 2021 - link

    Zen also has auto overclocking. The 5800X also gets pretty toasty. They're just more efficient at it.
  • drothgery - Wednesday, October 27, 2021 - link

    You do realize that's almost 100W less than Rocket Lake, right?
  • tamalero - Thursday, October 28, 2021 - link

    two wrongs do not make it a right. They have to aim to the competitors not their own worst crap.
  • Qasar - Thursday, October 28, 2021 - link

    " praising the new intel architecture " um if this was a new architecture, wouldnt intel of called this gen 1, not gen 12 ?
  • Mr Perfect - Wednesday, October 27, 2021 - link

    The fact that they're actually specing a turbo wattage is at least an improvement. The 11th gen only had data on base clock TDPs listed, with turbo wattage left as an exercise for the consumer.

    Here's hoping that the turbo TDPs are accurate, but I expect board OEMs will once again blow well past the "normal" turbos by default.
  • Wrs - Wednesday, October 27, 2021 - link

    It's looking like Intel's latest 10nm is comparable in efficiency to TSMC's N7P. Additional power density in that case should be viewed as headroom rather than a handicap. A 5900X only has ~240W headroom on ambient cooling; clearly the 12900K has more if Intel is already able to rate the package at 241W stock.
  • Spunjji - Thursday, October 28, 2021 - link

    "It's looking like Intel's latest 10nm is comparable in efficiency to TSMC's N7P"
    How?
  • Wrs - Thursday, October 28, 2021 - link

    Basic physical measures like transistor density/gate pitch, and efficiency comparisons of existing Intel 10nm products (Tiger Lake) with TSMC N7 products in the same power envelope (Ryzen 4000).
  • Spunjji - Friday, October 29, 2021 - link

    62MTr/mm^2 (Renoir) vs 53MTr/mm^2 (Ice Lake).
    Efficiency comparisons between those two chips aren't flattering, either - same goes for Tiger Lake and Cezanne. If you limit both to 15W, AMD win more often than not. Intel need 28W+ (4C) and 60W+ (8C) to open up a performance lead.
  • Wrs - Saturday, October 30, 2021 - link

    Comparing apples and oranges. Renoir was 8 Zen2 cores with HT disabled. Ice Lake-U was 4 Sunny Cove with HT enabled. Of course Renoir was better at MT, but slower at ST. It's a basic design tradeoff. An exaggerated example would be Intel shipping 12 Atom cores instead of 4 Willow Cove. Similar die area, but higher efficiency! Also more complaints about slowdowns.

    Also note Zen3's are on the same process as Zen 2, whereas ADL is on a newer process than I
  • Wrs - Thursday, October 28, 2021 - link

    And the power density difference is obvious and deliberate. Zen3 power is concentrated on the CCDs, ADL on the other hand is one piece of silicon over twice the size. Then measure the elevation difference between PCB and IHS on a Ryzen 5000, compare to reviewers doing the same on ADL. Thinner interface = better heat dissipation. That assumes other factors equal. Intel specifically talks about thin solder and thicker IHS, which is a further bonus because copper IHS has 5x the thermal conductivity of indium solder.
  • Oxford Guy - Thursday, October 28, 2021 - link

    Indium is also expensive and a depleting resource. Copper is rather more abundant.
  • mode_13h - Friday, October 29, 2021 - link

    Is it plausible that the thick IHS is there mostly as extra thermal mass to facilitate the PL2 turbo boost?

    I wonder if we'll ever see a CPU use a vapor chamber instead of solder + IHS. Just etch capillaries directly into the die surface and use some non-conductive liquid. Even solid copper is no match for the thermal conductivity of a vapor chamber!
  • Wrs - Tuesday, November 2, 2021 - link

    @mode_13h The thicker IHS is to reduce the bending moment, to compensate for the thinner solder leaving the die more fragile. Solder is a soft metal, but the chip is ceramic, cracks like a ceramic tile. And if they needed more thermal mass they'd fill in the void between IHS and substrate, where the chip isn't.

    On the vapor chamber idea... install heatsink wrong, chip explodes?
  • mode_13h - Friday, November 5, 2021 - link

    > if they needed more thermal mass they'd fill in the void between IHS and substrate,
    > where the chip isn't.

    Compared to the added volume of the IHS, I don't see that extra space as being very significant.

    > On the vapor chamber idea... install heatsink wrong, chip explodes?

    Huh? Vapor chambers are like heat pipes. Yes, it'd have to be mechanically strong, but I don't see why this approach would be significantly weaker than the CPUs we have today. And so what you're saying is that if you push down too hard on this, it'd fail just like if you push down too hard on a conventional CPUs. The risk is in breaking the substrate.

    Now, if you don't know about heatpipes, they're significantly *below* atmospheric pressure, at room temperature. So, while a CPU is cold, the real risk would be cracking it and creating a leak. The fluid would have to be something not terribly hazardous, in case it did leak out. However, I think phase-change cooling systems already commonly employ relatively benign fluids for similar purposes.
  • Spunjji - Friday, October 29, 2021 - link

    I'm aware of how these things work. As I said on your other comments, Intel are doing these things (die thinning, solder thinning) because they have to, thanks to the amount of power they need to dissipate to stay competitive. Up until Comet Lake they were slapping thermal paste under the IHS and calling it a day. I approve of these changes and I have little doubt that AMD will start implementing similar techniques as and when they need to (TDPs in general are on an upward trend).
  • GC2:CS - Wednesday, October 27, 2021 - link

    Well exciting times coming up will all the new CPU competition ramping up.

    I am jsut curious how can the “little core” drop off only about 20% perf while being 1/4 the size ?
    That is not a little core at all. More like lower clocked P core. If I remember corectly Apple little cores are only 1/3 of P cores and quite insignificant in MT workloads.
  • nandnandnand - Wednesday, October 27, 2021 - link

    Who knows. I think we still need official word or die shots to confirm the relative die size of the efficiency cores. They have 40% of the L2 cache and no hyper-threading which should make a difference.

    "A segment contains either a P-core or a set of four E-cores, due to their relative size and functionality. Each P-core has 1.25 MiB of private L2 cache, which a group of four E-cores has 2 MiB of shared L2 cache."
  • lmcd - Wednesday, October 27, 2021 - link

    You don't remember correctly at all. Apple's little cores are stupidly fast for little cores. Andrei flails in every Apple SoC review how stupid it is that there's no ARM licensed core answer to Apple's little cores.

    Intel probably roadmapped Alder Lake the minute they saw how performant Apple little cores were in even the iPhone 6S.

    Atom has been surprisingly good for a while. No need to make up conspiracies when you can buy a Jasper Lake SKU that confirms Intel Atom is far from slow.
  • name99 - Thursday, October 28, 2021 - link

    Apple's small cores are
    - about 1/3 the performance at
    - about 1/0th the power, net result being
    - same amount of computation takes about 1/3 the energy.

    The Intel cores appear (based on what's claimed) to be substantially faster -- BUT at the cost of substantially more power and thus net energy.
    If they are 70% of a P core but also use 70% of the power, that's net equal energy! No win!
    It won't be that bad, but if it's something like 70% of a P core at 35% of the power, that's still only half the net energy. Adequate, but not as good as Apple. My guess is we won't get as good as that, we'll land up at something like 50% of the power, so net 70% of the energy of a P core.

    (And of course you have to be honest in the accounting. Apple integrates the NoC speed, cache speeds, DRAM speed all ramped up or down in tandem with demand, so that if you're running only E cores it's your entire energy footprint that's reduced to a third. Will Intel drop the E-*core* energy quite a bit, but it makes no real difference because everything from the NoC to the L3 to the DRAM to the PCIe is burning just as much power as before?)

    Essentially Apple is optimizing for energy usage by the small cores, whereas Intel seems to be optimizing for something like "performance per area".
    That's not an utterly insane design point, but it's also not clear that it's a *great* design point. In essence, it keeps Intel on the same track as the past ten years or so -- prioritizing revenue issues over performance (broadly defined, to include things like energy and new functionality). And so it keeps Intel on track with the Intel faithful -- but does nothing to expand into new markets, or to persuade those who are close to giving up on Intel.

    Or to put it more bluntly, it allows Intel to ship a box that scores higher in Cinebench-MT at the same chip area -- but that's unlikely to provide an notably different, "wow", experience from its predecessor, either in energy use or "normal" (ie not highly-threaded) apps.

    Of course we'll see when the Anandtech review comes out. But this is what it looks like to me, as the salient difference between how Apple (and, just not as well, ARM) think of big vs little, compared to Intel.
  • nandnandnand - Thursday, October 28, 2021 - link

    "It won't be that bad, but if it's something like 70% of a P core at 35% of the power, that's still only half the net energy."

    I don't know how it will compare to Apple, but if it has a performance-per-area *and* a performance-per-watt advantage, it is a major improvement for x86. Especially as Intel iterates and puts 16 or 32 E-cores alongside 8 P-cores.

    Basically, Intel can continue to tinker with the P-cores to get the best possible single-threaded performance, knowing that 8 P-cores is enough for anyone™, but spamming many E-cores is Intel's path to more multi-threaded performance.

    Alder Lake can be considered a beta test. The benefits will really be felt when we see 40 cores, 48 threads (8+32) at the die space equivalent of 16 P-cores. The next node shrink after "Intel 7" will help keep power under control.
  • vogonpoetry - Wednesday, October 27, 2021 - link

    User-rewritable SPDs are a total game-changer for RAM overclockers. Many times I have wished for such a feature. As is on-the-fly power/frequency adjustment (though I wish we could change timings too).

    As for "Dynamic Memory Boost", doesnt Power Down Mode already do something similar currently? My DDR4 laptop memory frequency can be seen changing depending on workload.
  • Oxford Guy - Thursday, October 28, 2021 - link

    All overclocking is dead.
  • Oxford Guy - Thursday, October 28, 2021 - link

    I should have said: 'All user overclocking is dead'.

    Vendor-approved overclocking (i.e. going beyond JEDEC) is another matter.
  • Silver5urfer - Wednesday, October 27, 2021 - link

    On paper it looks okay. Staring with the Z690 chipset is a really deserved upgrade, lot of I/O plus RAID mode optimizations. AMD RAID is so bad, Level1Techs also showed how awful it was.

    STIM is interesting, given how 10900K and 11900K improved vastly with Delidding and LM. So that's a plus. Then the whole Win11 BS is annoying garbage. The WIn11 OS is horrible anti user anti desktop anti computing it reeks desperation to imitate Apple as an Ape. It looks ugly, has Win32 downgrades with integration to UWP, Taskbar downgrades, Awful explorer UI. It's outright unacceptable.

    Now the big part CPU and Price - Looks like Intel is pricing it wayy lower than AMD. For unknown reasons as Intel never does it, I find it disrupting. Also the CPU OC features are somewhat okay I was expecting lower clocks but looks like 5.1GHz but looking that the new PL1 system I do not have a problem at all since I want full performance now no more BS by GN and etc citing omg the power limits 125W must be kept on a damn K unlocked processor. But there were rumors on power consumption going past 350W like RKL once OCed that's the reason why Intel is going 8C max unlike Sapphire Rapids Xeon at 14C. DDR5 is also on it's new life not worth investing money into the DDR5 new adopter tax if DDR4 works which is what I'm curious about RKL Gear 1 4000Mhz is impossible. I wonder how this will fare.

    The leaked performance preview shows mediocre improvements, the ST is definitely a lead on the P cores, real cores. But the SMT / HT is really what I'm interested vs 10900K and Ryzen 5900X. RKL is also fast in ST but SMT was okay not great because 14nm backport.

    I'll be waiting to see how Intel executes this, not going to invest tbh because new chipset, new CPU design, Win11 and I want to run Windows 7. I'd rather settle for a 10900K on Z590. PCIe SSDs are not much of a value for me, they have no use beyond load times and boot times for my use case, MLC 860 Pro SATA SSD is way better, runs cool, long lasting as well.
  • Gothmoth - Wednesday, October 27, 2021 - link

    people who do raid without a dedicated PCIe RAID controller have no clue anyway.

    while most focus on performance i am waiting on performance per WATT figures.
  • vegemeister - Wednesday, October 27, 2021 - link

    Hardware RAID is a recipe for weird bugs and data loss, and provides no benefit over software RAID on top of the same controller running as a dumb HBA.

    Motherboard fake RAID is similarly pointless.
  • Gigaplex - Thursday, October 28, 2021 - link

    I'd rather do mdadm software RAID or use ZFS vs a PCIe RAID controller.
  • Makaveli - Wednesday, October 27, 2021 - link

    Can you post a link to this AMD Raid = bad?
  • Silver5urfer - Wednesday, October 27, 2021 - link

    Watch Level1Techs on RAID.
  • Spunjji - Thursday, October 28, 2021 - link

    All motherboard RAID is crap
  • Silver5urfer - Wednesday, October 27, 2021 - link

    Interesting aspect I completely forgot. Class 2 and 1 basically are AVX. Meaning Cinebench R23 AVX workload and Games so E cores won't even be used here, which are Class 3. How the gaming is even going to work on these.....Class 0 is where Intel says most. Meaning IDT at play for the general purpose compute.

    I want to see how this CPU behaves with all P cores only, does the power plane share anything with the E cores ? or how does it Impact with OC. This will be the most important question to Ian.
  • psychobriggsy - Thursday, October 28, 2021 - link

    ADL is not super-expensive because of the additional platform and DDR5 cost.
    12600K(F) is cheap because Intel have a lot of them due to defect density in Intel 7 process, but with so many cores it is easy to reclaim these dies as a lower product.
  • yeeeeman - Wednesday, October 27, 2021 - link

    amd needs a price drop
  • 5j3rul3 - Wednesday, October 27, 2021 - link

    New Epic!
  • Slash3 - Wednesday, October 27, 2021 - link

    "With Intel’s Z690 boards and 12th Gen Core Alder Lake processors, that changes. Much like processors and graphics have had idle states and turbo states for generations, memory now gets it as well."

    If I'm not mistaken (it's been a long night), Tiger Lake mobile CPUs can do this, too. You can see the memory frequency change from XMP to base and back in a utility like CPU-Z, depending on workload.
  • dwade123 - Wednesday, October 27, 2021 - link

    Intel wins!

    BTW... is Anandtech forum still full of retarded AMD mods censoring anything that's not pro-AMD?
  • lmcd - Wednesday, October 27, 2021 - link

    Your comments are more likely being censored for using language that doesn't belong in civilized society.
  • Hifihedgehog - Wednesday, October 27, 2021 - link

    Well, that rules you out, Pip... err... lmcd.
  • Spunjji - Thursday, October 28, 2021 - link

    Evidently not, given the presence of ghastly fanbots like you.
  • nicamarvin - Wednesday, October 27, 2021 - link

    These gaming numbers are with crippled Win11 for AMD, so right there the 15% is out of the windows and more like 5%, Amd 3D Cache will land when a mature W11 is out and they will take back the gaming lead by a good margin
  • Wrs - Wednesday, October 27, 2021 - link

    It is still an assumption that L3/L4 cache quantity would help that much for gaming. Remember increasing the cache size also increases latency. Ever notice the 5800x beating the 5950x in games? The 5800x has 32mb L3 on one CCD, the 5950x 64mb L3 split across two CCDs. There's good reason it's taking AMD a whole year to release 3d-cache on the same cores. It's a heterogeneous implementation and takes a lot of work and trial & error to minimize latencies.
  • nicamarvin - Wednesday, October 27, 2021 - link

    AMD said that 3D Cache would not increase latencies
  • Gigaplex - Thursday, October 28, 2021 - link

    "Remember increasing the cache size also increases latency."

    Not always. Going from L2 to L3 increases latency, but doubling the size of L3 doesn't increase latency.

    "The 5800x has 32mb L3 on one CCD, the 5950x 64mb L3 split across two CCDs."

    That's a NUMA type issue, not a cache size issue.
  • Spunjji - Thursday, October 28, 2021 - link

    "It is still an assumption that L3/L4 cache quantity would help that much for gaming"
    Not really. L4 on Broadwell helped a bunch, and this is tripling the L3 cache with little-to-no latency penalty. What remains to be seen is whether or not they can retain high clock speeds, and what the cost is.
  • nicamarvin - Wednesday, October 27, 2021 - link

    "Intel shows showing the 12900K being anywhere from slightly slower than the Ryzen 9 5950X, to being 30 percent faster. However, Intel does admit that these benchmarks were captured on Windows 11 before the performance patch for AMD CPUs was available, so the results aren’t as meaningful as they would have been had they tested the 5950X in its best performing mode, such as using Windows 10 or waiting for the patch to be available"
  • WaltC - Wednesday, October 27, 2021 - link

    I am very underwhelmed so far...but time will tell. I am, however, delighted to see Intel put out a new architecture at long last...;) Something tells me that seeing the most game performance possible will require *a lot* of optimization--so hopefully Intel will have good compilers ready to go! The article says that Z690 is a PCIe 4.0 chipset--is that right? Does that mean the Intel hasn't yet produced a PCIe5.0-supporting chipset? Didn't quite understand that one.

    Going to be interesting to see how this shapes out...;)
  • Spunjji - Thursday, October 28, 2021 - link

    Alder Lake has PCIe 5.0 lanes from the CPU - 16 of them - but there are no PCIe 5.0 lanes from the chipset, and the chipset connects to the CPU over a PCIe 4.0 equivalent link.
  • Duncan Macdonald - Wednesday, October 27, 2021 - link

    One stupid thing that Intel does - each new generation has a different socket and needs a new motherboard. AMD uses the same socket for multiple generations allowing upgrades by changing the CPU by itself.

    Given the problems that Windows 11 has on AMD systems at the moment, the best comparison is with one of these new Intel CPUs vs AMD CPUs running Windows 10. (As is the norm for Microsoft any new release of an OS from them needs at least 6 months of patches before it is reasonably stable and performs fairly well.)
  • Bik - Wednesday, October 27, 2021 - link

    Big part of windows 11 existence is for intel core 12.
  • tygrus - Wednesday, October 27, 2021 - link

    I can't remember a time when Windows was ever out of beta. They keep messing with it, users do most of the testing after releases/updates, more updates come months/years later to fix some but creates more bugs requiring more updates...
  • nandnandnand - Wednesday, October 27, 2021 - link

    Windows 10 was supposed to be the last version, but here we are.
  • Gigaplex - Thursday, October 28, 2021 - link

    It's usually every 2 generations that Intel changes socket, not 1.
  • Spunjji - Thursday, October 28, 2021 - link

    They often change chipset between socket changes as well, though
  • Oxford Guy - Thursday, October 28, 2021 - link

    'One stupid thing that Intel does - each new generation has a different socket and needs a new motherboard.'

    That's why this article passage is a bit droll:

    'However, in that time, we’ve pretty much kept the same socket design for mainstream processors. There hasn’t been much emphasis on changing the design itself for thermomechanical improvements in order to retain reuse and compatibility. There have been some minor changes here and there, such as substrate thinning, but nothing that substantial. The move to a new socket for Alder Lake now gives Intel that opportunity.'
  • Harry_Wild - Wednesday, October 27, 2021 - link

    Will, not one CPU is for lower PSU setups! 😱😕
  • Alistair - Wednesday, October 27, 2021 - link

    So first Intel charges too much for motherboards.

    In Canada a 5800x (nevermind the 5600x, i'm being generous) is $470, the motherboard Asus Tuf ATX is $130. $600 total.

    For Intel, the 12600k is $390 and the motherboard is a whopping $290, just for Prime, not Tuf. That's $680.

    For AMD ram is $200 for 32GB. For Intel DDR5 is a whopping $650.

    Total AMD: $800
    Total Intel for those benchmarks: $1330.

    Might as well buy the 5950x and motherboard and ram for less than the cheapest Intel CPU, Motherboard, and DDR5.
  • nandnandnand - Wednesday, October 27, 2021 - link

    Come to USA, we have cheaper bread and circuses.

    That DDR5 price looks like an aberration though. It's expected to be near +60%, not +225%.
  • Spunjji - Thursday, October 28, 2021 - link

    Here in the UK, 32GB of DDR5 (4800 C40, blank DIMMs) starts at £279. You can get 32GB of DDR4 (3600 C18, RGB) for £140.
  • maroon1 - Wednesday, October 27, 2021 - link

    Alder Lake support ddr4
  • Spunjji - Thursday, October 28, 2021 - link

    Cool, that's only $100 more expensive and you kill your chance of an upgrade to Raptor
  • Roy2002 - Thursday, October 28, 2021 - link

    Lowest price for Z690 motherboard is $180. Comparable to Z490 boards 2 years ago.
  • Roy2002 - Thursday, October 28, 2021 - link

    Plus, there are many boards support DDR4.
  • Alistair - Wednesday, October 27, 2021 - link

    To be honest Alder Lake is looking way worse than I expected. Pricing is ludicrously bad as you can't just buy a CPU, Intel is responsible for those high motherboard prices (they charge extra license fees for Z platforms).

    On top of it, the 19 percent claimed IPC is then only 13 percent in games. But Intel used a broken copy of Windows and outliers to make that claim.

    It actually loses in Tomb Raider? Ties in Crysis? What?!?!?!

    Terrible.
  • Alistair - Wednesday, October 27, 2021 - link

    Actual gaming advantage? We have to wait for next week, but looks like less than 10 percent on average for sure.
  • dwade123 - Wednesday, October 27, 2021 - link

    Angry AMD fanatic spotted. Let's face it. Zen 3 is now Slowzen 3 and SlowZen3D is still the same old Slowzen 3 cores with more useless caches glued to it. :)
  • Makaveli - Wednesday, October 27, 2021 - link

    lol this reminds me of someone that post on wccf is that you Silky?
  • Gigaplex - Thursday, October 28, 2021 - link

    If anyone is the fanatic it's you, using childish names like that.
  • Alistair - Thursday, October 28, 2021 - link

    no, i was excited to upgrade, but if I can't even beat AMD at Tomb Raider, why bother?
  • Spunjji - Thursday, October 28, 2021 - link

    P r o j e c t i o n
  • maroon1 - Wednesday, October 27, 2021 - link

    Implying that gaming is the only thing that matter when it comes to cpu

    12600k and 12700k should easily destroy 5600x and 5800x in productivity
  • Gigaplex - Thursday, October 28, 2021 - link

    Using twice the power, I'd hope it's twice as fast in productivity.
  • Alistair - Thursday, October 28, 2021 - link

    huh? Gaming is the only point of Intel. If you are doing productivity you get a 5950x.
  • Oxford Guy - Thursday, October 28, 2021 - link

    Not sure one of those is needed to run Excel.
  • Spunjji - Thursday, October 28, 2021 - link

    Not implying anything, it's just what he cares about apparently
  • Roy2002 - Thursday, October 28, 2021 - link

    This is huge gen to gen performance leap, you decided to ignore?
  • Bik - Thursday, October 28, 2021 - link

    "Over the 11900K, the 12900K is +50% faster in BlenderMT at 241W (vs 250W)"
    This roughly about the same performance as amd 3900x. The 5900x is about 10% faster. teh outlook of 16 cores intel vs 12 cores amd not looking great for team blue.
  • Bik - Thursday, October 28, 2021 - link

    not to mention 241w vs 142w
  • laduran - Thursday, October 28, 2021 - link

    This is an amazing generational leap in performance from Intel. I hope this performance leap survives the leap to mobile parts. How awesome would it be to have 20 thread Core i5 based laptop that can game as well as have all day battery life?
  • Alistair - Thursday, October 28, 2021 - link

    it is 16, not 20
  • Farfolomew - Thursday, October 28, 2021 - link

    I think Intel missed the bus on this. Apple has shown the way forward on future SoC designs: highly integrated and fast memory bandwidth, joint GPU/CPU memory access, and very large GPU die. The hybrid design is just a small part of what makes the M1 so powerful and efficient. I'm hoping AMD can spin up a design similar to their console SoCs that can compete with the M1.
  • Farfolomew - Thursday, October 28, 2021 - link

    Boat, bus, whatever...I knew that didn't sound right LOL
  • Farfolomew - Thursday, October 28, 2021 - link

    Why didn't Intel use this opportunity to rebrand their Core line? They're all the way up to 12th gen, surely it's time to come up with something new? Brand matters, sure, but if this is as revolutionary as Intel claims, (over 10 years since such a radical change has occurred, per the beginning of this article) then come up with a new name for cryin' out loud! At the very least, keep the Core moniker, but call it something different like maybe Core h(ybrid)X. ie, Core h5, Core h7, Core h9. Do something! Sheesh
  • Hifihedgehog - Thursday, October 28, 2021 - link

    Intel Fusion, maybe? Core is pretty unoriginal when you think about it.
  • kwohlt - Thursday, October 28, 2021 - link

    My understanding is that each "Core" from Intel has been an iteration, and that the underlying Microarchitecture originates from the original Core 2 line from 2006, and that an entirely new, from the ground-up, architecture is expected to replace Core in 2025 (Royal Core Project)
  • Farfolomew - Thursday, October 28, 2021 - link

    That would be awesome, I hope that's true. This current Core line dates back to the Pentium Pro, iterated to the Pentium III mobile, then turned into Core 2 Duo, Core iX, and now here. So yeah, over 20 years of the same underlying architecture. Maybe this Royal Core Project is a departure from x86? ARM with powerful emulated x86 would be delightful.
  • Silver5urfer - Thursday, October 28, 2021 - link

    What nonsense is that ? ARM already disposed off 32Bit blocks from their new 2022 designs and you think magically that ARM BS will emulate x86-x64 ? I want what you are smoking. I think it's that new Apple miracle crack only found on interwebs.
  • iranterres - Thursday, October 28, 2021 - link

    These "new" core performance designations are the perfect strategy when you know you have severe power efficiency issues, day one.

    LOL @ 125w
  • Freeb!rd - Thursday, October 28, 2021 - link

    The problem with the MSRP listed for these processors is that at MicroCenter all the Intel SKUs are going for $40-60 HIGHER than MSRP, while the AMD SKUs are going for $40-60 LOWER than MSRP.
  • nandnandnand - Thursday, October 28, 2021 - link

    AMD Zen 3 prices were above MSRP for a long time. Maybe give this a week?

    And certainly don't go pre-ordering these before the reviews are out.
  • WaltC - Thursday, October 28, 2021 - link

    Newegg today selling pre-orders of the 12900 for ~$649, IIRC...;) Hard to believe they'd sell any--but judging by some posts here, I can't rule it out...;)

    Z690 mboards are all pre-order, too.
  • Flying Aardvark - Thursday, October 28, 2021 - link

    Looks like AMD will continue sucking on it. Intel never lost the AAA (or overall) gaming crown. That will continue on. One day AMD will overtake in something that's not CSGo. Just don't hold your breath.
  • Qasar - Thursday, October 28, 2021 - link

    you should hold your breath, as you are wrong, most reviews show and state that intel lost everything but the gaming crown with Zen 2 and lost the gaming crown with zen 3, the reviews show and state this.
  • Flying Aardvark - Saturday, October 30, 2021 - link

    I disproved you on that lie with benchmarks. You're just the typical fanboy that doesn't know what he's looking at and believes everyone that parrots falsehoods just because Zen3 does win in some titles. Good for AMD, but they didn't take the gaming crown. Never once did AMD beat Intel in the majority of titles.
  • Qasar - Sunday, October 31, 2021 - link

    " I disproved you on that lie with benchmarks " no you didnt, you disproved nothing.
    " You're just the typical fanboy that doesn't know what he's looking at and believes everyone that parrots falsehoods just because Zen3 does win in some titles " hello pot, meet kettle. you base your whole argument on a geomean ranking, when IF you were to look at the reviews of each cpu them selves, you would see quite the opposite.
  • Flying Aardvark - Monday, November 1, 2021 - link

    I proved that you don't know what you're talking about because you're not up to date on MCE/ABT and the impact of that on early reviews. My argument is based on the truth. The average consists of individual titles. You have a pretty dumb argument.
  • Qasar - Monday, November 1, 2021 - link

    as do you. point is
  • Flying Aardvark - Monday, November 1, 2021 - link

    Seems like you're out of things to say and just angry now. Enjoy being stupid.
  • Qasar - Monday, November 1, 2021 - link

    nope. just no point argueing with someone that is a blind intel shill who wont look at individual tests to see that they are wrong. and that those reviews showed and stayed intel lost the gaming crown with zen 3.
  • Spunjji - Friday, October 29, 2021 - link

    I'll never understand the point of lying this far into a comments section
  • Qasar - Friday, October 29, 2021 - link

    simple, intel fanboy maybe ?
  • Flying Aardvark - Saturday, October 30, 2021 - link

    It's you AMD fanboys that lie. I'm the one revealing them. Just be happy with your hardware, no one cares if have a 2nd place gaming CPU. The gap has definitely closed so at least it's no longer humiliating to run AMD, but they are still buggier and less friendly with memory than Intel to this day. I know, I have a 5900X rig alongside an 11900K. I'll take the i9 every single day over the Ryzen. It's just a better overall machine.
  • Qasar - Sunday, October 31, 2021 - link

    no, you are just blind the the truth, and you have the 2nd place gaming cpu, individual reviews show this, look them up, like AT bench for example. i agree, it is no longer humiliating for amd, its now humiliating for intel :-)
    " but they are still buggier and less friendly with memory than Intel to this day. I know, I have a 5900X rig alongside an 11900K. " must be just you then, i have been using a 3900x for about 1.5 years now, issue free the whole time, and just picked up a 5900x, and so far, was a drop in replacement, and still no issues. " I'll take the i9 every single day over the Ryzen. It's just a better overall machine." that sucks power from the wall :-)
  • Flying Aardvark - Monday, November 1, 2021 - link

    Best AAA gaming chip = humiliation. You're worried about power consumption for the few hours you'll be gaming.. really grasping for straws just like every other fanboy.
  • Qasar - Monday, November 1, 2021 - link

    just like you your self are doing
  • Flying Aardvark - Monday, November 1, 2021 - link

    Nope, I just told you and every other delusional fool the truth. And I'll have the 12900K soon too, enjoy your potato.
  • Oxford Guy - Thursday, October 28, 2021 - link

    'It got to a point where Intel’s power consumption under turbo became a bit of a meme, and enthusiasts got annoyed that Intel buried this information away.'

    Intel's use of TDP was worse than buried information. It was deceptive. Glad to see it has belatedly been axed.
  • Oxford Guy - Thursday, October 28, 2021 - link

    I strikes me as ridiculous that people are being told that it's a good thing for a large amount of adequate hardware to be orphaned, via Windows 11's very aggressive Apple-like support policy — just so Alder Lake can be supported better.

    Is that the responsibility of consumers? To prop up Intel's latest design by being forced off a lot of useful hardware? It seems to me that if Intel wants Windows to cater to its CPUs it's its responsibility to get that, rather than dumping the problem onto consumers' wallets.

    Windows 10 is not going to have security patches after a few years. So, please don't try to excuse this with the argument that people can simply stay on 10. They can't. The Internet is at the heart of computing now. It is no longer 1984.
  • kwohlt - Friday, October 29, 2021 - link

    "...just so Alder Lake can be supported better."
    Windows 11 aggressive cut-off is based on a lot of things, such as VBS support, Spectre-Meltdown hardware mitigations, etc. It's not due to accommodating Alder Lake whatsoever.
  • GeoffreyA - Friday, October 29, 2021 - link

    It is nonsense but more related to trying to push increased security in an arbitrary way. Truth is, W11 runs happily on a Pentium 4. Microsoft, the cat is out of the bag that the requirements are a sham.

    Anyhow, I suspect W11's coming out in quite a rough state was due to Alder Lake's release.
  • Oxford Guy - Friday, October 29, 2021 - link

    If MS continues to copy Apple it will implement hard locks to prevent it from running on anything but its arbitrary list.

    People should consider the incrementalist implications of MS’ radical change from being the OS of long-term hardware compatibility to being Apple Jr.
  • GeoffreyA - Saturday, October 30, 2021 - link

    Let's hope they don't follow in Apple's footsteps. Microsoft has a history of relenting when their choices haven't been sober. I understand they're trying to push increased, military-grade security; but when you're cutting out 90% of the world's computers, it doesn't seem sensible. As in life, moderation is best. W11 doesn't have a purpose, and though I haven't used it yet, reminds me of ME. It hasn't reached critical mass of change to warrant a departure from 10, which works flawlessly, and whose minimal approach fades away into the background, much like XP's did.

    As for your sentiment touching on consumer passiveness, you're right. If people got together and boycotted these tech products, when they're bad, companies will have to give in. Unfortunately, we just accept rubbish from them, and they get away with worse and worse. Turning the tables round, the famed entitlement of consumers also deserves some comment.
  • Oxford Guy - Sunday, October 31, 2021 - link

    'Turning the tables round, the famed entitlement of consumers also deserves some comment.'

    What entitlement?

    Have you shopped at a WalMart recently? One goes into the store, no one is available to check you out. When the machine (which was filthy — to the point where the attendant doesn't even know where the spray bottle is) goes haywire, the attendant comes over and wasted a great deal more of your time trying to get the broken computer to work. The attendant argues with you about the problem you can see with your own eyes and proceeds to overcharge you. If you have a problem with that you'll have to come to the store again when a manager is willing to be at the front area.

    That doesn't go into the big Orwellian televisions on every self-scan in some stores, the managers who lie to your face about product pricing and disappear as soon as you try to check out the alleged sale items, and people who are tasked with blocking your exit and making harassing comments.

    WalMart has been forced on consumers and it is leading the way in anti-entitlement.

    The now-legendary passivity in 'geek' tech is to be seen in the passivity concerning all of these degradations of the shopping experience.
  • GeoffreyA - Sunday, October 31, 2021 - link

    I don't even know where to start, and agree the shopping experience is anything but ideal. Not being very fond of shopping myself, I try to limit how much I go into the shops. I would say, all customers want is good service, friendliness, honouring what prices are listed, etc.

    Here in South Africa, Game and Makro are our Walmart analogues, and we've still got cashiers at the tills. Incidentally, Walmart acquired a majority stake in Massmart, the company that owns all these shops.
  • Oxford Guy - Thursday, October 28, 2021 - link

    'Perhaps it’s time for some new words.'

    Shakespeare coined many.
  • Oxford Guy - Thursday, October 28, 2021 - link

    Perhaps I didn't read carefully enough but I didn't see anything about ECC in the DRAM section. I saw a bunch of complex new tech for RAM (turbo and such).

    Maybe it's strange to think that prioritizing data safety should come before these other things.
  • mode_13h - Friday, October 29, 2021 - link

    Why would they mention ECC? This is a consumer CPU. Intel consumer CPUs generally don't support ECC memory, with but a few exceptions (i3's usually do, and select other low-end SKUs over the years).

    The Xeon E-series equivalent will almost surely support ECC memory*.

    * DDR5 supports internal ECC, but the datapath still doesn't have ECC bits by default, since that requires additional motherboard traces which translates into precious $0.001's (sarcasm).
  • Oxford Guy - Friday, October 29, 2021 - link

    ‘Why would they mention ECC? This is a consumer CPU.’

    Agreed. Consumers don’t have much use for RAM that reduces data loss/corruption. They need new stuff like turbo in their RAM. Priorities.
  • mode_13h - Saturday, October 30, 2021 - link

    I'm just pointing out that Intel's market segmentation practices essentially rule out inclusion of ECC in a mainstream platform. That could always change, but I've seen no indication that it would.

    What you *want* to happen is a separate matter.
  • Oxford Guy - Sunday, October 31, 2021 - link

    'What you *want* to happen is a separate matter.'

    Glad you've realized that. Now, back to the subject of ECC not being mentioned...
  • mode_13h - Saturday, October 30, 2021 - link

    FWIW, I wish ECC support were more widespread. I always use ECC memory, when it's an option, but it really restricts ones choice of CPU & motherboard.
  • Oxford Guy - Sunday, October 31, 2021 - link

    'I wish ECC support were more widespread.'

    Then your conduct here should reflect that.
  • mode_13h - Sunday, October 31, 2021 - link

    > Then your conduct here should reflect that.

    FFS, dude. *Your* conduct should reflect better internet skills. Your original post indicated what could be genuine uncertainty over the lack of ECC being addressed. I replied to clarify the matter to you and anyone else who might be interested in the subject. And this is the MF thanks I get? Geez, you're welcome but you can keep your snide remarks.

    Sure, there was a note of sarcasm in your post, but it's hard to know just how much might've been sincere. I'm sorry for giving you the benefit of the doubt. You should know by now that text isn't an effective medium for implicit sarcasm. Don't act like this is my fault.
  • Oxford Guy - Friday, November 5, 2021 - link

    I can’t be bothered to read your latest defensive rant. Fact is that when you attack someone’s post for advocating for ECC RAM you don’t get to simultaneously claim you’re an ECC RAM advocate.

    Figure out what you actually think prior to posting.
  • mode_13h - Friday, November 5, 2021 - link

    > when you attack someone’s post for advocating for ECC RAM

    Try not to get your ego so tied up in your posts that you can't distinguish a factual counterpoint from attacking your motives.

    I never attacked the value of ECC. I pointed out that Intel doesn't put ECC in their mainstream consumer CPUs. Your post was worded as if you were unaware of this fact. If you are trying to question Intel's priorities, then make it clear.

    > Figure out what you actually think prior to posting.

    I have never changed my position on this matter. Not for 23 years, which is the first time I owned a machine with parity memory.

    And stop shifting blame. Be better.
  • mode_13h - Friday, October 29, 2021 - link

    I don't see why the DDR5 era poses such a challenge in describing DIMM configurations. The difference between DDR4 and DDR5 is that the latter has a pair of channels, rather than a single channel. So, instead of talking about "Dimms Per Channel" (DPC), simply use the phrase "Dimms Per Channel-Pair (DPCP). Problem solved.
  • mode_13h - Friday, October 29, 2021 - link

    "we’ve seen Intel over the years transition from a soldered down heatspreader, to liquid metal, to basic thermal paste (because saving 0.1 cents means a lot across 70m CPUs)"

    First, which CPUs used liquid metal?

    And I simply can't believe the figure of thermal paste saving them only $0.001 per CPU. It seems to me that the delta of soldering the dies would be more than that. Furthermore, that amount over 70 M CPUs is only $70k, which I think is not enough to justify the performance impact. For a long time, Intel said it was competing against itself, which meant that a smaller generational uplift meant fewer upgrades. So, that $70k of savings could easily be overcome by a small boost in sales that could come with an additional % in performance.

    So, is there any credible information that the savings Intel got by using thermal paste was REALLY that small?
  • Samus - Friday, October 29, 2021 - link

    Prescott is BACK BABY! Just look at that pipeline and those power draws!
  • yeeeeman - Friday, October 29, 2021 - link

    You comparisons at various price points are a good idea, but wrong, because 12600K will fight with 5800X performance wise, hence its efficiency will be judged compared to the 5800X, not with the 5600X, which obviously more efficient than the 5800X.
  • Spunjji - Friday, October 29, 2021 - link

    This remains to be seen - Intel rarely go low on price and high on performance. I'll be happy if they've changed that pattern!
  • Carmen00 - Friday, October 29, 2021 - link

    Surprised that nobody's talking about the inherent scheduling problems with efficiency+performance cores and desktop workloads. This is NOT a solved problem and is, in fact, very far away from good general-purpose solutions. Your phone/tablet shows you one app at a time, which goes some way towards masking the issues. On a general-purpose desktop, the efficiency+performance split has never been successfully solved, as far as I am aware. You read it right - NEVER! (I welcome links to any peer-reviewed theoretical CS research that shows the opposite.)

    In the interim, it seems to me that Intel has hitched its wagon to a problem that is theoretically unsolvable and generally unapproachable at the chip level. Scheduling with homogenous cores is unsolvable too, but an easier problem to attack. Heterogenous cores add another layer of difficulty and if Intel's approach is really just breaking things into priority classes ... oh, my. Good luck to them on real-world performance. They'll need it.

    I can see why they've done it. They have the tech and they're trying to make the most of their existing tech investment. That doesn't mean it's a good idea. It is relatively easy to make schedulers behave pathologically (e.g. flipping between E/P constantly, or burning power for nothing) during normal app execution and we see this on phones already. Bringing that mess to desktops ... yeah. Not a great idea.
  • kwohlt - Friday, October 29, 2021 - link

    "I welcome links to any peer-reviewed theoretical CS research that shows the opposite"

    That's not how this works. YOU are making the claim that heterogenous scheduling has never been solved on a general-purpose desktop - the onus would be on your to provide proof of this.

    "Scheduling with homogenous cores is unsolvable too"
    So if scheduling with heterogenous and homogenous cores is both unsolvable, then what point are you trying to make?

    Has apple not demonstrated functional scheduling across a heterogenous architecture with M1?
    And what does "solved" look like? Because if your definition of solved is "no inefficiencies or loss", then that's not a realistic expectation. Heterogenous architecture simply needs to provide more benefit than not - a goal of zero overheard of inefficiency is unrealistic.

    As long as efficiency and performance gain outpaces the inefficiencies, it's a success. Consider a scenario of 4 efficiency cores occupying the same physical die space, thermal constrains, and power consumption of 4 performant cores - Surely an 8 + 8 design would offer better performance than a 10 + 0 design, when the e cores can offer performance greater than 50% of the P core.

    Consider this: a 12600 will be 6 + 0. A 12600K will be 6+4. If we downclock 12600K to match P core frequency of 12600, we can directly measure the benefit of the 4 e cores.
    If we disable 4 e cores in the 12700K so it is only 8P cores, and compare that to the 6+4 of the 12600K, if the 12600K is more performant, we can directly show that 6+4 was better than 7+0 in this scenario.
  • Carmen00 - Wednesday, November 3, 2021 - link

    Sure, take a look at this for a good recent overview: https://dl.acm.org/doi/pdf/10.1145/3387110 . (honestly, though, I think you could have found that one for yourself... the research is not hidden!)

    Again, best of luck to Intel on it; deep learning models or no, they have a risk appetite that I don't share. Apple's success is due, in no small part, to the fact that it controls the entirety of the hardware and software stack. Intel has no such ability. My prediction is that you will have users whose computers sometimes simply start burning power for "no reason", and this can't be replicated in a lab environment; and you will have users whose computers are sometimes just slow, and again, this can't be replicated easily. The result will likely be an intermittently poor user experience. I wouldn't risk buying such a machine myself, and I certainly wouldn't get it for grandma or the kids.
  • mode_13h - Saturday, October 30, 2021 - link

    Good scheduling requires an element of prediction. In the article, they mention Intel took many measurements of many different workloads, in order to train a deep learning model to recognize and classify different usage patterns.
  • PedroCBC - Saturday, October 30, 2021 - link

    In the WAN show yesterday Linus said that some DRMs will not work with Alder Lake, most of then old ones but Denuvo also said that it will have some problems
  • iranterres - Friday, October 29, 2021 - link

    Having "efficiency" cores in modern desktop CPU are irrelevant. For laptops is another story.
  • iranterres - Friday, October 29, 2021 - link

    *is irrelevant
  • nandnandnand - Friday, October 29, 2021 - link

    False. It boosts multicore performance per die area.

    Intel could have given you 10 big cores instead of 8 big, 8 small. And that would have been a worse chip.
  • Bik - Saturday, October 30, 2021 - link

    Let's check your argument. If 4 small cores > 1 big core while using same die area, how about just one big core for single thread performance and give us an addition of 24 more small cores? Clearly Intel (and Apple) not doing that, rather they reduce number of small cores (m1 pro, core i5 alderlake) instead. There clearly a catch, we'll know when review is out.
  • Bik - Saturday, October 30, 2021 - link

    7*4=28 cores. my math sucks.
  • Wrs - Saturday, October 30, 2021 - link

    For 1P 28E they'd probably run into fabric/coherency limits. If you have 30 nonlocking threads then that could be a rather efficient design. Let's see if we have that in a modern computer... *stares at the 10.5k CUDA cores on an RTX card*
  • nandnandnand - Saturday, October 30, 2021 - link

    Simple. You can have more than one task running at a time that can't be parallelized and will only respond to greater single-thread performance, and one or more applications running that can use multiple threads but not scale to benefit from dozens or hundreds of cores.

    Games will typically be using up to 8 cores in the upcoming generation because the PS5 and XSX have 8 cores and will be optimized to get the most out of 8. So Intel puts 8 big cores in the flagship but stops there. If the game can also use some of the small cores, great.

    Also, we already know that next year's Raptor Lake will have up to 8 big cores and 16 small cores, and it looks like Intel will go to 8 big cores and 32 small cores after that. They can tweak to improve IPC across all the cores, but the small cores will be getting a greater proportion of the die area going foward. Apple has a different strategy, if their leaks are correct.
  • ThinkingReviewer200 - Friday, October 29, 2021 - link

    Thanks for the article :)

    For the second page of the review, the sentence, "For performance, Intel lists a single P-core as ~19% better than a core in Rocket Lake 11th Gen" may have an inaccuracy. This percentage should perhaps be ~14% for according to the graph image 128% - 112% is 16% then 16%/112% is ~14% depending on the perspective, in this case perhaps the 11th gen. However, check this logic and math for these are percents inside percents calculations.
  • mode_13h - Saturday, October 30, 2021 - link

    They're not additive. It's easier to see, if you drop the percentages and just call a Rocket Lake core 1.12x as fast as a Comet Lake core. Then, Alder Lake-P is 1.28x as fast as comet lake. So, Alder Lake-P should be 1.28/1.12 as fast as a Rocket Lake core which is 1.143 or about 14%.
  • mode_13h - Saturday, October 30, 2021 - link

    Ah, but if you look closely at the chart on page 2, it says "fixed-frequency". So, that's talking about essentially "instructions per clock". However, the stats on the last page are including clock speed differences.

    So, I guess that's how they get up to 19%? It would take only a clock speed increase of about 4.3% to makeup the difference. It's still not easy to see where they find that, but if both cores throttle back and Alder Lake's P core manages to hold a higher clock speed, then it's not hard to see how we could reach a 4.3% difference.
  • tech4fun - Friday, October 29, 2021 - link

    I just canceled my Netflix account. Fanboi battle royale!
  • nandnandnand - Friday, October 29, 2021 - link

    What took you so long?
  • Kendog52404 - Sunday, October 31, 2021 - link

    Does Alder Lake support DDR4-3000 memory? Or, does it need to be 3200? I haven't bought memory for about 5 years, and it's DDR4-3000 that I had been planning to bring over from the old board. If I have to buy new memory, regardless, that changes things, and I'll need to take a serious look at DDR5.
  • mode_13h - Sunday, October 31, 2021 - link

    In my experience, they list the upper limit but you can always go lower. The thing you really want to check is voltage. The next thing to pay attention to is the rank.

    Your best bet is to look at the list of approved memory for the motherboard you buy. Even if you don't buy memory specifically listed there, it'll at least give you some idea of what combinations of frequencies and voltages are supported. Your next best option is to use the compatibility apps on memory manufacturers' websites, which will tell you which of their memory is supported for a specific motherboard.
  • Meteor2 - Monday, November 1, 2021 - link

    Zen 3 has the highest IPC and perf/Watt in the x86 world. Will be interesting to see if Intel can close those gaps when the reviews come out.
  • vlado08 - Tuesday, November 2, 2021 - link

    https://ark.intel.com/content/www/us/en/ark/produc...

    What do you think means:
    Graphics output HDMI 2.1
    Max resolution (HDMI) 4096 × 2160 @ 60 Hz

    When I saw HDMI 2.1 in some mother boards I thought 4k @ 120 Hz but it seem it is not.
  • cheshirster - Monday, January 17, 2022 - link

    Windows 11 Does It Best

    How it works in real life on 12700K
    1) Minimized task goes E-cores only (4 threads out of 20)
    2) Maximized task uses all cores
    3) If a window of the maximized task is covered with another window, it goes to E-cores only again.

Log in

Don't have an account? Sign up now