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

Page 319 - 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
805
1,389
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).



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:
Reactions: richardllewis_01
Jul 27, 2020
17,491
11,280
106

My very "bad math" prediction: Zen 4 could be 108% in the relative performance graph. Raptor Lake may be 110%. It's gonna be close. Anything more than this would be a pleasant surprise.
 

Timmah!

Golden Member
Jul 24, 2010
1,444
684
136
Then you'd have people on the west coast bitching. There is no perfect answer. It's not like 1am is all that bad. I'd stay up for it.

Europe is surely bigger market than California, so if i was AMD and deciding whose bitching is more important to me, i would consider that.

I Will most likely stay up for it, i usually go to sleep past 1AM anyway.
 

Schmide

Diamond Member
Mar 7, 2002
5,588
719
126
Europe is surely bigger market than California, so if i was AMD and deciding whose bitching is more important to me, i would consider that.

I Will most likely stay up for it, i usually go to sleep past 1AM anyway.

Maybe in GDP, but in processors, Lawrence Livermore National Laboratory probably buys more processors than some small EU countries.

Edit: remember if Cali was a country it would be #5 on the list.
 
Last edited:

Carfax83

Diamond Member
Nov 1, 2010
6,841
1,536
136
So if 5.8ghz is the single core boost, all core boost is probably going to be 4.7/4.8ghz.

I would be shocked if the 7950x could hit 4.7/4.8ghz all core boost and remain at the 170w threshold.
 

eek2121

Diamond Member
Aug 2, 2005
3,027
4,215
136
I wouldn't expect Raphael to be much more efficient than Raptor Lake at stock, but we'll see. They're pushing it pretty hard with the presumed 230W PPT.

Quotes like yours will be gems when the new chips drop. Intel is (self admittedly) on an inferior process. There is no possible way Intel can reach the perf/watt of Zen 3, much less Zen 4 at stock settings.

I wish this forum had ban bets...
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,722
14,750
136
Quotes like yours will be gems when the new chips drop. Intel is (self admittedly) on an inferior process. There is no possible way Intel can reach the perf/watt of Zen 3, much less Zen 4 at stock settings.

I wish this forum had ban bets...
Not only that, there is no 230 PPT, thats the socket max, not any chips. Idiot.
 
Reactions: Tlh97 and Drazick

turtile

Senior member
Aug 19, 2014
618
296
136
So if 5.8ghz is the single core boost, all core boost is probably going to be 4.7/4.8ghz.

I would be shocked if the 7950x could hit 4.7/4.8ghz all core boost and remain at the 170w threshold.

It makes sense to me. If the process clocks a lot better, it's probably a lot more efficient at a much higher clock. I bet the last Ghz takes a lot more power.
 

v.strix

Junior Member
Aug 25, 2022
8
18
41
GB5 does not care about Memory bandwidth. Thats why an OC Alder Lake Celeron with 2C/2T can match a 8C/16T Alder Lake.
Eh... When a 2C/2T matches a 8C/16T in a multi-thread test you really find it convincing that's all because it's overclocked, rather than because its memory is very similar?

I looked at a few random samples of both Intel and AMD desktops and there's nearly linear correlation between memory frequency and MT crypto score.

I'll repeat it again. MT AES-XTS subtest of geekbench is a pathetic microbenchmark that degenerates into memory bandwidth measuring tool on any reasonable hardware.
I fully expect That 7950X double the crypto score of the 13900K
See above.
It will go up compared to where it was on zen 3 due to DDR5, but that's about it. ST variant should see an improvement though.

And i very much doubt any other subtests support AVX-512 at all...
 
Reactions: MadRat

Carfax83

Diamond Member
Nov 1, 2010
6,841
1,536
136
It makes sense to me. If the process clocks a lot better, it's probably a lot more efficient at a much higher clock. I bet the last Ghz takes a lot more power.

For heavy loads like rendering or encoding though which tap all 16 cores to near 100% usage? And some codecs have AVX-512 optimizations. It will be interesting to see how that affects the 7950x's power and frequency scaling.
 
Reactions: Tlh97 and Drazick

v.strix

Junior Member
Aug 25, 2022
8
18
41
Not only that, there is no 230 PPT, thats the socket max, not any chips. Idiot.
Sockets don't consume power, chips do.

So, that means you guys don't believe those rumors that the top of the upcoming lineup will have 170W TDP?

And that AMD:
  • Ordered OEMs to implement those power requirements on MBs only for it to be a dead weight laying there unused until some years later.
  • Decided to artificially limit the performance of their upcoming halo CPUs. Just because. Even though all the power delivery infrastructure is in place.
Like, when was the last time something like this happened?
 

Thunder 57

Platinum Member
Aug 19, 2007
2,793
4,067
136
Europe is surely bigger market than California, so if i was AMD and deciding whose bitching is more important to me, i would consider that.

I Will most likely stay up for it, i usually go to sleep past 1AM anyway.

It's an American company. They will prioritize the US when it comes to things like this.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,722
14,750
136
They have nearly identical PPT (13900K 250W, 7950X 230) and 7950X has 50W higher TDP (though that is just a paper spec, under load both will hit their PPT), what are you referring to?
You have not read anything ? 170 watt is the PPT of the 7950x. For the 13900k we have no official word, but rumors put it at 350 watt. Since you and all the other Intel supporters don't believe in facts, lets just see what happens when its all in the reviewers hands. I put my money in the fact that the 7950x will consume 100 watts lets than the 13900k. Thats full SUSTAINED power, not a quick benchmark.

Damn I am sick of trolls.
 

Vope45

Member
Oct 4, 2020
114
168
86
Sockets don't consume power, chips do.

So, that means you guys don't believe those rumors that the top of the upcoming lineup will have 170W TDP?

And that AMD:
  • Ordered OEMs to implement those power requirements on MBs only for it to be a dead weight laying there unused until some years later.
  • Decided to artificially limit the performance of their upcoming halo CPUs. Just because. Even though all the power delivery infrastructure is in place.
Like, when was the last time something like this happened?

Robert Hallock said the 7950x was running well under 170w in the demo with all cores running well above 5Ghz. It's literally less than a day away, wanna bet something interesting?
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,101
136
170 watt is the PPT of the 7950x. For the 13900k we have no official word, but rumors put it at 350 watt
Since you clearly haven't been following rumors, the 13900k is rumored around 250W, and 170W is the top TDP, not PPT, of AM5. So maybe stop calling people trolls for keeping track better than you? Then again, given that you consider even reading source links to be trolling...
 
Reactions: Carfax83

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,722
14,750
136
Since you clearly haven't been following rumors, the 13900k is rumored around 250W, and 170W is the top TDP, not PPT, of AM5. So maybe stop calling people trolls for keeping track better than you? Then again, given that you consider even reading source links to be trolling...
Well, thats not what I read. I saw 350. And you and others keep equating socket AM5=7950x, and thats crap. Socket AM5 may or may not be the same, but 170 watt is what I have read. See the post above yours dude.

trolls again.
 
Reactions: Tlh97 and Drazick

Exist50

Platinum Member
Aug 18, 2016
2,452
3,101
136
Well, thats not what I read. I saw 350. And you and others keep equating socket AM5=7950x, and thats crap. Socket AM5 may or may not be the same, but 170 watt is what I have read. See the post above yours dude.

trolls again.
Well this matches perfectly. You consider informing oneself (yes, that includes reading) to be trolling, so you never bother.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,722
14,750
136
If your claims are on topic, then me pointing out how ludicrous they are surely is as well. If anything, you can't seem to go a single comment without ranting about Intel.
ludicrous ? Like claiming that what AMD says is true ???? You need to learn English as well. Or at least how to read.

Edit: Oh, wait, you believe everything Intel says is true, so thats where your stupidity comes from.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |