Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)

Page 317 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,572
14,525
136
No, we are at zen 5 being 40% faster than Zen 4 unless it is 10% faster or 5% slower than Zen 4 (in IPC or single thread or multi-thread or spec or some mixture therein) and it will both be crushed by arrowlake and will crush arrowlake (maybe even at the same time) in both the same and different workloads depending on the time of day.
I believe it was qualified, it was one specific spec benchmark by a server ship that was over 40% higher. If I am wrong, sorry.
 
Jul 27, 2020
16,340
10,352
106
IF it is correct, AMD had Granite Ridge samples running at over 6 GHz, ST boost. Not by much, but still above 6 GHz, stable.
Is it possibly because Zen 5 is using some better thermal interface between the die and the heatspreader? Better than the one used in Zen 4?

I bet Lisa Su is carrying a sample around in her purse and it's like a useless trinket to her, just to wow clients and close friends. Who's up for devising a plan to rummage through her purse and get a photo of the sample when she goes to the little ladies' room at some conference or event or whatever?
 

uzzi38

Platinum Member
Oct 16, 2019
2,635
5,984
146
Im pretty sure I have the same info as Kepler and Adroc, and potentially Uzzi.

IF it is correct, AMD had Granite Ridge samples running at over 6 GHz, ST boost. Not by much, but still above 6 GHz, stable.
I probably do, but I will point out that to my understanding the number you're thinking of is analogous to the 5.85GHz FMax on Raphael rather than the 5.7GHz rated boost clock on the box. Just to be clear about that number.

Anyway I'm going back to shutting up on Zen 5 again, bye.
 

Gideon

Golden Member
Nov 27, 2007
1,646
3,712
136
I am sorry, this is not specifically about Zen 5, but AMD literally did this just a few years ago. Apple did it with the M1 vs both x86 and ARM. Why is it so hard to believe? Because Intel isn't doing it? Intel has been stuck and spinning their wheels to get anywhere. Don't get me wrong, they ARE working on some stuff, but they were far too reliant on node improvements and that absolutely kicked them in the rear end.

Qualcomm, AMD, and Apple have dropped or are dropping chips between this year and next year that are going to bring large improvements to the table. It doesn't happen every year, yes, but it happens frequently. Graviton, for example.

I made a very specific claim. No-one in the x86 market has in the last 20 years done anything close to ~40% IPC gain in a single generation, especially from an already cut throat competitive staring point (zen 4). Let's tackle your claims one-by-one:
  • Zen 1 doesn't really count as Bulldozer was terrible (Unlike Zen 4). It was an insane leap, but still only about matched Skylake in IPC (in integer at least) while still clocking lower.
  • Core 2 Duo doesn't count as while it beat P4 soundly, it was nowhere near 40% uplift from it's mobile predecessor (Yonah)
I never mentioned ARM but since you brought it up:
  • M1 was rad, but nowhere near 40% from A13 and while it offered insane IPC gains vs Comet Lake and even Tiger Lake (50%+), it also clocked about as much lower. It was incredible as far as efficiency goes but not all-out performance. Don't get me wrong, it was a huge achievement, but a very iterative stride not "one big jump":

  • Graviton 1 was a very uncompetitive starting point for G2. I guess G3 is actually pretty close to 40%, but that's still multiple ARM generations A-76 (N1) -> A-77 -> V-1 (enlarged A-78) .
  • Qualcomm's Snapdragon X is an insane IPC uplift indeed but it's from a Coretex A78 (which already has 3 succeeding cores A710, A715, A720) and isn't really beating the competition in absolut 1T performance.

Look, I'm all for Zen 5 to be great, in fact I'm really rooting for it (just keeping healthy skepticism until there is at least some proof thrown to the plebs like us) , but if the rumors are true, it really is an unprecedented achievement in a looooong time.

Even if it's "just" ~40% total ST uplift (with those rumored 6 GHz Clocks) instead of ~40% IPC, it will outperform everything out there in raw 1T performance. M1 didn't do that, Snapdragon X certainly won't do that. I guess Core 2 Extreme was the last time something similar happened in workloads like SPEC (but half of that was a Zen 4 level clock-speed uplift rom Yonah, not IPC).

LATE EDIT: spell check
 
Last edited:

gdansk

Platinum Member
Feb 8, 2011
2,123
2,629
136
One can also consider that Zen 4 is actually quite far behind the competition in IPC. It's a modern speed demon Pentium 4 or Bulldozer when you compare to Apple's Everest. Would 30-40% increase in IPC even exceed it? The IPC claim itself is not too unbelievable but combination of high IPC and high clocks would be something new.
 

Gideon

Golden Member
Nov 27, 2007
1,646
3,712
136
I remember the Rv770 speculation and the 800 shader leak, that seemed unbelievable yet it is exactly what we got.
This was glorious! Particularily because it was done with the help of AMD's engineers (not long after the purchase of ATI) that hand designed the new ALUs (compared to the syntesized ones on the RV 670) allowing the 320 -> 800 shader unit uplift on the same 55 nm node!

RV 770 really was an exiting chip (and a good reminder how damn good journalism Anandtech used to be). I highly suggest reading the article from that time from this site:


And that was just half a year after the release! I really wish, we'd get similar articles from current GPU developments today.

One can also consider that Zen 4 is actually quite far behind the competition in IPC. It's a modern speed demon Pentium 4 or Bulldozer when you compare to Apple's Everest. Would 30-40% increase in IPC even exceed it? The IPC claim itself is not too unbelievable but combination of high IPC and high clocks would be something new.

I don't know. M3 Pro is 30% faster than M1 Pro, while clocking ... about 30% higher (4.05GHz vs 3.2GHz) with a major node upgrade and power draw increase. Apple's cores would indeed be similar beasts if they clocked to 5.5 GHz ... but they don't.
 
Last edited:

Goop_reformed

Member
Sep 23, 2023
153
182
76
I probably do, but I will point out that to my understanding the number you're thinking of is analogous to the 5.85GHz FMax on Raphael rather than the 5.7GHz rated boost clock on the box. Just to be clear about that number.

Anyway I'm going back to shutting up on Zen 5 again, bye.
People often forget these test samples are pushed always to the limits, hence the "test samples"
 
Last edited:

H433x0n

Senior member
Mar 15, 2023
895
965
96
Really wish we’d just get a proper benchmark leak. For reference we got Zen 4 benchmark data in July 2022, roughly 2.5 months before proper release date. If Zen 5 is supposed to launch in June that’d mean we should expect leaked benchmarks any day now.
 

gdansk

Platinum Member
Feb 8, 2011
2,123
2,629
136
Really wish we’d just get a proper benchmark leak. For reference we got Zen 4 benchmark data in July 2022, roughly 2.5 months before proper release date. If Zen 5 is supposed to launch in June that’d mean we should expect leaked benchmarks any day now.
I'm not sure leaked benchmarks are as likely when the platform - AM5 - is already tested.
We'll know better from how vague (or not) AMD is at Computex.
 

Gideon

Golden Member
Nov 27, 2007
1,646
3,712
136
Really wish we’d just get a proper benchmark leak. For reference we got Zen 4 benchmark data in July 2022, roughly 2.5 months before proper release date. If Zen 5 is supposed to launch in June that’d mean we should expect leaked benchmarks any day now.
Isn’t it supposed to be announced at Computex? Release might still be some time after. Zen 4 announcement was on August 29 but launch on September 27, 2022
 

Glo.

Diamond Member
Apr 25, 2015
5,711
4,559
136
Shipping manifests of Granite Ridge and Zen 5 based processors.


So it begins...
 
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/    |