Intel Announces 12th Gen Core Processors Specifications aka Alder Lake

Horus-Anhur

2021-10-27 22:11

schmidtbag:

Even if the results weren't before the patch, they're still reasonable. The 12900K has higher boost clocks, is on DDR5, and Intel has typically had an edge when it comes to latency. So on a technical standpoint, the 12900K ought to have an edge in gaming regardless of the patch. Granted, it's a moot point because we're talking a few FPS when it's already in the hundreds. I'm sure the 5950X is an overall more powerful CPU, but that's just simply not going to show up in games. Frankly, the 12600KF ought to be more than good enough for nearly all games for several years.

I haven't seen game benchmarks with and without the Zen patch on Win11. So I can't tell how much that Ryzen 5950X is losing in performance. I it's very likely to close the gap a bit.

#5959513

Undying

2021-10-27 22:16

Horus-Anhur:

I haven't seen game benchmarks with and without the Zen patch on Win11. So I can't tell how much that Ryzen 5950X is losing in performance. I it's very likely to close the gap a bit.

If you consider initial pre-patch ryzen was loosing up to 15% performance. Thats a very big deal and makes intel performance claims irrelevant. [youtube=wFJ4x9FiLQI]

#5959516

kanenas

2021-10-27 22:19

Horus-Anhur:

I haven't seen game benchmarks with and without the Zen patch on Win11. So I can't tell how much that Ryzen 5950X is losing in performance. I it's very likely to close the gap a bit.

depending on the game I would say 3-10%

#5959519

Horus-Anhur

2021-10-27 22:25

Undying:

If you consider initial pre-patch ryzen was loosing up to 15% performance. Thats a very big deal and makes intel performance claims irrelevant.

This will put the Zen3 very close to Alder Lake. At least in games.

#5959524

mackintosh

2021-10-27 22:39

I'll be surprised if the patch really makes that much difference in everyday benchmarks. I think their gaming results and their lead over Zen3 will hold. That said, on average less than 20% performance gain from a new architecture is disappointing. Gone it seems are the days of Athlon, Conroe, Nehalem, or Sandy Bridge. This isn't even Skylake.

#5959526

ViperAnaf

2021-10-27 22:56

AMD patch might have effected the latency of the L3 in benchmarks but burly cost 1fps in gaming benchmarks so cut the BS fanboys... https://images.hothardware.com/contentimages/newsitem/56521/content/ryzen-windows-11-perf-farcry.png https://images.hothardware.com/contentimages/newsitem/56521/content/ryzen-windows-11-perf-shadow-of-war.png

#5959527

Serotonin

2021-10-27 23:18

cucaulay malkin:

you bought dlss

Not with a 2080. I bought a gpu that everyone mocked me for as unnecessary because it was "identical" to the 1080Ti. I bought ray tracing. Dlss was introduced later, an added feature r benefit to the card. 30xx users bought dlss as it was an already available and known feature upon purchase.

#5959528

Horus-Anhur

2021-10-27 23:19

mackintosh:

I'll be surprised if the patch really makes that much difference in everyday benchmarks. I think their gaming results and their lead over Zen3 will hold. That said, on average less than 20% performance gain from a new architecture is disappointing. Gone it seems are the days of Athlon, Conroe, Nehalem, or Sandy Bridge. This isn't even Skylake.

The average of perf improvements over Zen3, that Intel presented on it's own slides, for games, is just 12%. And this is without the Zen patches for Win11. But other applications are probably gain more, I suppose.

#5959534

tsunami231

2021-10-28 00:20

tty8k:

If those prices will follow in shops, AMD looks like a total rip off, which probably was for about 1 year. According to performance difference we had between 10k and 5k series 1 year ago which is similar to this but viceversa, Intel should charge $400 for a 12600k (not to mention 5800x is still 450-500 in shops lol) Who is/was greedy now eh?

dont give intel more reason to have price premium

#5959536

Agonist

2021-10-28 01:05

ViperAnaf:

AMD patch might have effected the latency of the L3 in benchmarks but burly cost 1fps in gaming benchmarks so cut the BS fanboys... https://images.hothardware.com/contentimages/newsitem/56521/content/ryzen-windows-11-perf-farcry.png https://images.hothardware.com/contentimages/newsitem/56521/content/ryzen-windows-11-perf-shadow-of-war.png

Over react much. Jesus. So stating things makes someone a fanboy. Block list you will go.

#5959538

mackintosh

2021-10-28 01:34

@tty8k They're both greedy as sin. AMD priced Zen3 at a premium because Intel had nothing to respond with and they saw the market was going to absorb the increase due to a unique set of circumstances. If 11th gen didn't end up being such a lackluster mediocre product, AMD would have been forced to slash its pricing after a couple of months. Unfortunately, Intel let them get away with it for an entire year. Make no mistake, Intel would have had done the exact same thing - and they will, given the opportunity. I imagine with ADL they want to win back some of that market share they lost, which is why they decided to price it aggressively. That, and a ~20% performance increase isn't exactly the revolution people were hoping for. Gotta entice them somehow. Either way, this is a (small) net gain for us. At the very least AMD will heavily discount Zen3 in the run up to Zen3D - so those who prefer team red will get a good deal on an outgoing platform. Team blue otoh will get a shiny new socket, with a shinier CPU for - arguably - a very good price given today's market. Personally, both corporations can kiss my hairy arse, I'm waiting for either Raptor Lake or Zen4.

#5959540

user1

2021-10-28 01:58

tty8k:

I don't think that crap DDR5 4400 is doing it any favor to be honest. Could be better with a low CL DDR4

Actually it might be ok, the fact that they are running the ddr5 well below spec (4800) implies they may be forcing gear 1 mode ( full rate imc), so it might be closer to what you would expect from highend ddr4 than actual ddr5 ironically.

#5959555

kanenas

2021-10-28 06:29

ViperAnaf:

AMD patch might have effected the latency of the L3 in benchmarks but burly cost 1fps in gaming benchmarks so cut the BS fanboys... https://images.hothardware.com/contentimages/newsitem/56521/content/ryzen-windows-11-perf-farcry.png https://images.hothardware.com/contentimages/newsitem/56521/content/ryzen-windows-11-perf-shadow-of-war.png

BS?

#5959557

Reddoguk

2021-10-28 07:13

No one here mentioned these new Power States or new way of reporting TDP which has me questioning whether all these chips sit @ 125w @ idle or what? Why report all cpus at 125w minimum even the ones without a GPU? Sounds a bit dodgy to me. I hope i'm wrong but did they have to really boost up TDPs to be a minimum of 125w because if they ran at a lower clock and idled lower you'd think they'd mention that but having two TDPs is strange.

#5959559

fry178

2021-10-28 07:20

125w is stndby 😀

#5959578

jose2016

2021-10-28 09:16

Yes, 125w in PL1 and 228w in PL2. Waiting for a review to see all the good and bad of these processors.

#5959579

RED.Misfit

2021-10-28 09:23

Undying:

How can you tell? Either way another reason to wait for Hilbert review.

The Windows' version is reported in the appendice, i've read them on techpowerup who posted the full list of documents that are not under NDA anymore. At the time of testing, there was no fix avaiable even as beta-testing (1st october). And of course, they didn't take 5 seconds to write "performance may vary for Ryzen because of L3 cache latency bug and prefered core tech bug mechanism" or something like that.

#5959588

mackintosh

2021-10-28 10:17

AMD's sudden onset emasculation and apparent inability to force Microsoft to fix the issue sooner is hardly Intel's problem, tbh - and it's not up to Intel to determine the severity of this issue, either. They tested with what was available on that day.

#5959590

lukas_1987_dion

2021-10-28 10:32

Intel know its a bug with W11 for Ryzen CPU's and still decided to go with the tests, so convenient.. 🙄 Imagine this, if there were no bug in W11 with Ryzen, Intel would show what, that its 5% faster than x5950? In true those CPU's are not much faster than last gen Intel and Ryzen CPU's, sometimes they are even slower due to the weird architecture and incompatibility with DRM's. I'm skipping this gen too. It would be cool to see tests on both W10 and W11. P.S. PCI-E 5 and DDR5 are the most exciting features from this gen to me. Next year should be interesting 😀

#5959614

Fediuld

2021-10-28 12:42

mackintosh:

AMD's sudden onset emasculation and apparent inability to force Microsoft to fix the issue sooner is hardly Intel's problem, tbh - and it's not up to Intel to determine the severity of this issue, either. They tested with what was available on that day.

Microsoft released W11 and immediately the issue was found, took 3 weeks for MS to fix it which is understandable even if bit slow. If Intel got a W11 the day was release run the test and prepared the presentation is not AMDs fault. They knew the issue existed and run with it. They could have used pre-release build from mid September where this issue didn't existed. It only existed with the release build somehow!! I am not the only one who things something fishy happening between Intel-MS. Do not forget MS took 7 months to fix the scheduler in 2017. Still has issues with Ryzen CPUs when playing game, you still required to use the Enterprise edition on Threadrippers to work properly! Yet on Linux magically everything is gone. Having games like X4 Foundations and X Rebith running happily on 16 core CPU stretching it's legs and perf difference shows when game progresses after 100 hours where the world is full of ships doing their business.