Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)

Page 525 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Vattila

Senior member
Oct 22, 2004
807
1,411
136
Except for the details about the improvements in the microarchitecture, we now know pretty well what to expect with Zen 3.

The leaked presentation by AMD Senior Manager Martin Hilgeman shows that EPYC 3 "Milan" will, as promised and expected, reuse the current platform (SP3), and the system architecture and packaging looks to be the same, with the same 9-die chiplet design and the same maximum core and thread-count (no SMT-4, contrary to rumour). The biggest change revealed so far is the enlargement of the compute complex from 4 cores to 8 cores, all sharing a larger L3 cache ("32+ MB", likely to double to 64 MB, I think).

Hilgeman's slides did also show that EPYC 4 "Genoa" is in the definition phase (or was at the time of the presentation in September, at least), and will come with a new platform (SP5), with new memory support (likely DDR5).

Untitled2.png


What else do you think we will see with Zen 4? PCI-Express 5 support? Increased core-count? 4-way SMT? New packaging (interposer, 2.5D, 3D)? Integrated memory on package (HBM)?

Vote in the poll and share your thoughts! :)
 
Last edited:
  • Like
Reactions: richardllewis_01

JoeRambo

Golden Member
Jun 13, 2013
1,814
2,105
136
I don't think there is need for sensationalism over a CPU uArch on either side.
These things takes years of design and implementation, would lose hair fairly quick getting riled up over such long spans of design time.

Fully agreed. Over long time I have especially noticed that such sensationalism always starts when when companies are neck to neck and not when one of them is clearly in the lead. We already had Dr. Sharikou era of "Intel Bankrupt in Q3 07", for sure we will survive this latest bout of sensationalism.

Still, out of all clearly pro-AMD people, i respect only @Markfw as he actually went out and bought 12700K to compare it in his very special usage case. Others just blabber around with Skylake in mind and Alder Lake equals 12900K SKU and 666W of power use to them. Usually they own 3rd hand Z1 2600, but love talking about servers and imminent Intel's doom ( where of course every single sale is top SKU of 64C ).

Welcome to Anandtech Forums, where people love to discuss hardware they don't own! :)
 

NostaSeronx

Diamond Member
Sep 18, 2011
3,705
1,231
136
Zen4c has so far only been attributed to lower-density higher performance node.
"The Zen 4c chiplet, according to AMD, is built on an HPC variant of TSMC N5."

First iteration is on N5HPC (virtually built on N5HPC), then production iteration is on a tailored N4X variant(physically built on N4X).

HPC:
Wider Cell CPP
Wider Cell Height
Wider Metal Pitch
Wider Metal Via

HPC != UHD/Improved HD

The biggest processor from a competitor in the Zen4c category is 128-cores at 5.7+ GHz within 950W TDP.
 
Last edited:

NostaSeronx

Diamond Member
Sep 18, 2011
3,705
1,231
136
That is Wrong. Zen4C is HIGH Density not lower..
Zen4c is HPC, of which HPC has:
Wider poly pitch
Wider cell height
Wider metal pitch
Wider metal vias

The high density isn't from the libs, but rather from the macros/tiles.
Library = Less Dense
Architecture = More Dense

Same ISA;
Zen4 = Full 512-bit rate
Zen4c = Half 512-bit rate
 

NostaSeronx

Diamond Member
Sep 18, 2011
3,705
1,231
136
What are you on about? Is that sarcasm? How do you even cool something at 950W?
Tachyum 128-core = >5.7 GHz at 950W TDP
It is targeting the same field of Cloud Native, but without the need of discrete Cloud-native GPUs(SIMD/SIMT accelerators) or TPUs(AI accelerators). TDP is basically CPU+GPU+AI in one package.
128-cores + 1 MB L2 per, no discrete L3, instead uses inactive L2 as L3.

128-core >5.7 GHz is liquid cooled <=== 950W
128-core >4 GHz is air cooled. <== 700W/600W

ARM hasn't released their cloud-native cores yet. However, they did drop some hints in a forum that they will be targeting higher clocks than 3 GHz with N2.

None of the high-end cloud-native processors are targeting low-clocks, none of them.

N5HPC => Higher frequency from reduced capacitance and resistance on both FEOL/BEOL with wider structures.
Zen4c => Smaller area from cut-down core.

Less area means less wire delay, and less stage delay, etc. Cut-down reduces total dynamic leakage meaning something that takes 230W PPT to achieve might only need less than 115W PPT to achieve.
 
Last edited:
  • Wow
Reactions: igor_kavinski

NostaSeronx

Diamond Member
Sep 18, 2011
3,705
1,231
136
Tachym is a scam lol, it's a Theranos of chip designers.
It is just one of the known alternatives to Intel's monopoly of high frequency instances.

AmpereOneHF
AWS GravitonHF (High Frequency Neoverse cloud-native)
Alibaba Cloud T-headHF
etc.

The direction of Cloud-native development is towards High Frequency. Bergamo is launching in a swarm of 4~6 GHz cloud-native solutions.
Zen4 = 200W-400W
Zen4c = 320W-400W <-- 120W+ for 32 more smaller cores, yeah... Bergamo isn't going to be slow.

~Genoa lowest available TDP for 96-core ~200Ws (12x CCD @ ~16.7W)
~Bergamo lowest available TDP for 128-core ~320Ws (8x CCDs @ ~40W)

In everything I find, AMD is targeting the high-end of Cloud-native, and not the entry-end. This means AMD needs less supply to get a ROI and thus net profit.
 
Last edited:
  • Wow
Reactions: Grazick

Hans Gruber

Platinum Member
Dec 23, 2006
2,302
1,216
136
The good run of AMD is almost over. Enjoy Zen 4. It's not that AMD is failing. It's that Intel is improving generationally at a much faster rate. I am not saying AMD is returning to the decade of darkness and nothingness in CPU's. When Intel brings out Intel 4 aka. 7nm. That will bring Intel inline with AMD for power use. AMD can't keep up with Intel's IPC from generation to generation.

Send the hate. Intel was on 14nm for more than 6 years. Now they are on 10nm and after Raptor Lake they will be on 7nm. Ryzen had the fabrication advantage since Zen 2 over Intel. I should note that the power use for Intel to get the performance has not been practical. I think that will continue to some degree with Raptor Lake.

I will have a Zen 4 build of some kind within the next year.
 
  • Like
  • Haha
Reactions: Tlh97 and pakotlar

pakotlar

Senior member
Aug 22, 2003
731
187
116
Yes we all want 50% more IPC and 32 cores. But be reasonable. They can only diverge so much from existing designs without delaying their launch or taking a big risk. If things slip they are eaten by competitors who did make incremental improvements.

Then we’re in agreement. I would have been thrilled with the 8-10% IPC and 24 cores.

They’re delivering a rumoured 35% at +65W (+60%) when using 6000 MT DDR5 vs 3600 MT DDR4 on Zen 3, on a new full node (5nm). This is much worse than Zen 3 delivered without a new node or new memory generation (+19% IPC on the same 7nm). Yeah, MT is better than Zen 3, but it’s essentially an overclocked part + expensive DDR5 that gets us there.

Am I unhappy with this? Not really, it’s probably still better than Raptor Lake, and is pushing Intel to do better. Is it the weakest generational improvement since Bulldozer? Yep.

The exciting changes for me are AM5 (DDR5, PCIE5) and AVX512 support. The actual core is ok, but not nearly as exciting as exciting as Genoa and Threadripper Pro. It’s clear that they’re focusing their energy on server and workstation markets, hence all the core increases, 4+ channel memory is restricted to parts that costs thousands. Slowly leaving behind regular consumers. No one should be thrilled or defensive about that.
 
  • Like
Reactions: Timmah!

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
26,129
15,274
136
Aren't you the one who keeps insisting that we should wait for the official reviews from reputable sites? :p
Early data is not totally reputable, but his statement was to throw out the results. They make sense, in that in most cases genoa beats Milan. And as for the above comment. Golden cove <> sapphire rapids just the same as 5950x core <> Milan core.. Due to being a server, they certainly can be clocked lower. You Intel supporters should not comment on Zen 4 if all your feedback is negative. Wait for official reviews, and talk about them in Intel threads, otherwise it looks like you are troilling AMD threads.

And speaking of abandoning all reason, I say its you that has lost all reason.
 
Last edited:
  • Like
Reactions: Drazick

itsmydamnation

Platinum Member
Feb 6, 2011
2,920
3,545
136
It's certainly a larger gap than that on average. But regardless, the point is the same. The leak posits that Golden Cove is far weaker, but we know how it performs.
You have factored in the slow ass L3 in SPR compared to desktop ring when making that assessment right ?

actually has anyone done GC testing with the ring/ uncore clocked low to see its impacts ?
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
Excuse me. FIRST, Power. SR will have 6 watts or less per core, MAX. 12900k for the golden cores has around 29 watts per core. That is such a large difference that the frequency will be WAY different, and thus the performance. And due to the number of cores and design (mesh, ring, etc) there will also be a lot of differences. Do I have to spell this out ??
You can go read any number of reviews to know how Golden Cove performs at various power limits. Just because you refuse to acknowledge it doesn't mean the data isn't known. Now enough humoring this insanity.
 
  • Like
Reactions: Henry swagger

inf64

Diamond Member
Mar 11, 2011
3,864
4,546
136
Just to give more context to gaming performance, here is how things stand right now (on average @ 720P; Computerbase data ):


1659601533981.png


For 7000 series ( 8C/16T and up) to beat Alder Lake KS (~5.5Ghz boost part), they need to perform roughly 10% better than 5950X/5900X/5800X.
I find it very hard to believe that Ryzen 7000, with IPC and massive all core boost clocks cannot achieve ~10% better gaming performance on average Vs Ryzen 5000 series. It should easily beat 5000 series by at least 15% in games, basically beating the 5800X3D as well. This all goes for vanilla version of Zen4 without the Vcache.
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
So the people for whom the short duration performance is more important are trolls?
Remember, caring about anything more than power-limited, embarrassingly parallel HPC workloads is trolling, but claiming that Zen 3 is 4x as efficient as Golden Cove is merely an unbiased contribution.
 
  • Like
Reactions: Zucker2k

reb0rn

Senior member
Dec 31, 2009
238
73
101
Dream on.
Why, the TDP of raptor and zen4 is almost identical, why would you not expect similar power use for top models?
How you think they managed to push 5.5Ghz with multicore load where 5950x for me drop frequency way under 4Ghz with ~140w use, sure they must match intel speed and will push power use to the brink same as intel

You here dram a lies and fud a lot, ask yourself why non leak show power use for zen4
 

tamz_msc

Diamond Member
Jan 5, 2017
3,865
3,729
136
No offense to adam, but comparing anything to his highly tweaked and cherry picked CPU is crap.
Uh, both adam's 12900KS and the 7950X (and presumably the 7600X) were using 6000 MT/s DDR5, if I'm not mistaken. So if anything such a comparison is more apples to apples than pulling random scores from the Geekbench database.
 

Hans Gruber

Platinum Member
Dec 23, 2006
2,302
1,216
136
So 6000mhz for DDR5 right out the box maxes out the 1:1 fabric clock on Zen 4? DDR5 is only barely a year old. Does that mean DDR6 is right around the corner?
 

mikk

Diamond Member
May 15, 2012
4,243
2,297
136
So you are arguing about nothing….I see.

Zen 4 is far more efficient than Raptor Lake or Alder Lake. AMD raised power limits, which raised TDP, in order to unlock multicore performance. There is nothing wrong with this. Intel has high power limits because they are using a design that would not otherwise be competitive.

Just because of you, I am putting a 7950x in a mini ITX build.


It's barely faster with a full node advantage. Power consumption not much difference. Look above, only 10W difference. It seems like Zen 4 needs water cooling, if not performance drops. This might be a serious issue, even most reviewer using a traditional air cooler. Raptor Lake is much easier to cool down.
 

Borealis7

Platinum Member
Oct 19, 2006
2,901
205
106
The MT ratio is shockingly high. 21.55!
for comparison 5950x ALL cores OC 5.1Ghz only has 19.86 MT ratio:
stock 17.58
MT ratio seems like a very irrelevant and made-up metric, notice that neither Intel nor AMD use or cite it as a metric for anything.
the reason is the MT ratio just scales with the number of threads, each thread's frequency, the IPC of the architecture and most of all it is very workload dependent (the level of parallelism of the task you are performing). a CPU can have a MT ratio of 20 in one task and 15 in another...
you could compare the same task performed across many CPUs and measure the MT ratio, but the result would be not be a measure of efficiency unless you are comparing CPUs with the same number of threads and at the same frequency PER THREAD (or logical core). looking at the end result, and measuring performance/Watt is the industry standard and basically the only thing that matters because that translates directly into money.

I am just a humble B.Sc CompSci, but that is my view on the subject.
 
Last edited:

RnR_au

Platinum Member
Jun 6, 2021
2,036
4,939
106
That's actually a great change if they consistently stick to it.
Disagree.

The 'year' has no place being at the front. And why a '7' for 2022? Clearly to suggest some connection to Ryzen 7000 on the desktop. Its just marketing confusion for the masses.

Thankfully I don't have hordes of family members who buys computing gear before asking me 'why is this...' so the change won't affect me. Just saddened that marketing has won over engineering at AMD.
 

Just Benching

Banned
Sep 3, 2022
307
156
76
There is no way Intel can match AMD's power efficiency on Zen 4. That is using the new 5nm process. Intel is improving efficiency with Raptor Lake but they need the 7 nm process to make significant gains in efficiency. That will come next year or in 2024. Remember, Alder Lake is the 1st 10nm processor. Everything before that was 14nm.
Really? I think only the 7950x will be decent in efficiency compared to Raptorlake, the rest of the lineup will get beaten - significantly.

Same wattage the 7600x has no chance to win in efficiency against the 13600k or the 13400, and the same goes for the rest of lineup.


Also both the 13700T and the 13900T will be the most efficient cpus by far out of the box. There will be no contest.
 

Zepp

Member
May 18, 2019
171
163
116
I think Phoenix will be a good show case for x86 CPUs. RDNA3 will shine if rumors about focus on area and power efficiency are true.
Zen4 is also quite area efficient, on top of that the N4 node should give it a slight efficiency edge vs vanilla desktop Zen4. AIE could be interesting but probably not as interesting as it would have been on cellphone SoCs.

Next two years of Mobile products from AMD will be formidable,
Phoenix --> Zen4+RDNA3+AIE on N4.
Strix Point --> Zen5 + RDNA3+ and AIE on N3
It does seem AMD will remain unchallenged in terms of efficiency in x86 mobile space.
there is rumor that Phoenix will use RDNA2

 
  • Like
Reactions: ftt