• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

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

Page 658 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
And no, as @igor_kavinski wrote, I won't tell what companies I work or worked for 😀
Like @SarahKerrigan , you misattributed someone else's post to me but it's OK, I feel pretty flattered that mine is the first name that pops up in the head of hardcore CPU lovers 😛

 
I don't believe its a retail CPU, or other respected websites would have reviews too. Could be an ES. You can relink for me if you want, I did not see it.
I haven't watched the video either. It might be a retail sample though. Respected reviewers don't have reports out yet because AMD most certainly made them agree to a publication embargo, which ends at or near retail availability of the product. (And respected reviewers require more time anyway. The usual state of affairs has been that there is not much time between reception of review samples, reception of quasi-final BIOS, and end of publication embargo.)

Regardless if it is an ES or retail sample, the current owner of it is most likely not allowed to publish performance data or to give it to somebody who publishes performance data.
 
Ok so is the UCLK: 3600 and FCLK: 2400 on that Italian guy zen5 or
Did he just run some wierd 7200mts/ MCLK:3600 UFCLK:1800 FCLK:2400 combo ?
 
Ok so is the UCLK: 3600 and FCLK: 2400 on that Italian guy zen5 or
Did he just run some wierd 7200mts/ MCLK:3600 UFCLK:1800 FCLK:2400 combo ?
His answer to one of the comments at least makes it sound like he tested different configs and this was the best one.
Screenshot_20240723_203948_YouTube.jpg
 
Like @SarahKerrigan , you misattributed someone else's post to me but it's OK, I feel pretty flattered that mine is the first name that pops up in the head of hardcore CPU lovers 😛

My apologies to you and @FlameTail
 
Above 5.35ghz fmax (stock) together with 2x L1, ~1.4x L2 and ~1.15x L3 cache bandwidth
Ontop of 7ns L3 and the core IPC gains going from Z4 to Z5

What do you think the final gaming performance uplift will be vs the 7800x3d ? :innocent:

I'll stick with my assumption that vanilla Zen 5 is ~10% faster than vanilla Zen 4 in games, so I think that plus clock advantage over 7800X3D which would be 14% or so.

Theres a small chance that the L3 will allow the Zen 5 core to stretch its legs more than the Zen 4 core could, but thats pure speculation.
 
I would be surprised if he was able to get UCLK going at 3600. I'm expecting it was 2:1 mode at 7200/1800/2400
Highest IF on Zen4 is 2200 at like 1.35VSOC right ? And hes doing 2400IF. Wouldnt be unreasonable to think that max uclk is now 3600 up from 3300(highest iv seen).
Saw igor's lab comparison for 1:2 on zen4 its from 2022 on rtx3090 the difference is ~1-5% regresion for 1:2 vs 1:1
 

Attachments

  • Assetto-Corsa-Competizione-FPS-1440p-2.png
    Assetto-Corsa-Competizione-FPS-1440p-2.png
    81.4 KB · Views: 36
Highest IF on Zen4 is 2200 at like 1.35VSOC right ? And hes doing 2400IF. Wouldnt be unreasonable to think that max uclk is now 3600 up from 3300(highest iv seen).
Saw igor's lab comparison for 1:2 on zen4 its from 2022 on rtx3090 the difference is ~1-5% regresion for 1:2 vs 1:1
Personally I would suspect the entire setup is potentially unstable. Then again, there are very few people I have any faith in for tuning this kind of thing.

Zen 3 didn't see any real substantial increase in FCLK or UCLK over Zen 2 on the same IOD, maturity helped very little.
 
Last edited:
I would be surprised if he was able to get UCLK going at 3600. I'm expecting it was 2:1 mode at 7200/1800/2400

Highest IF on Zen4 is 2200 at like 1.35VSOC right ? And hes doing 2400IF. Wouldnt be unreasonable to think that max uclk is now 3600 up from 3300(highest iv seen).
Saw igor's lab comparison for 1:2 on zen4 its from 2022 on rtx3090 the difference is ~1-5% regresion for 1:2 vs 1:1
Have he actally said he was running 2400 FCLK ?
I very much doubt so.. More likely 1800 or 2000 i would say

*edit*
Found the comment
1721767202423.png

But i still don't believe he was running 2400mhz FCLK 🤣
 
Last edited:
It's not. Warhammer 3 battles are pretty insane.
Dunno about that.

Even 11 years after the first 8 core CPUs in game consoles it's still quite common to find newish games with bad multi threading and most of the heavy lifting dumped on only 1-2 threads, one of which contains the UI/main rendering.

Even UE5 does not address it well off the shelf.

Thankfully one of the items in UE5's "forward looking" roadmap is a big MT overhaul:

UE5_MT_01.jpg
 
Dunno about that.

Even 11 years after the first 8 core CPUs in game consoles it's still quite common to find newish games with bad multi threading and most of the heavy lifting dumped on only 1-2 threads, one of which contains the UI/main rendering.

Even UE5 does not address it well off the shelf.

Thankfully one of the items in UE5's "forward looking" roadmap is a big MT overhaul:
Pretty sure about it because I'm a Total War veteran. The campaign map is gigantic and the average battle usually involves up to 5000 units all moving at once. Double that with Skaven / Vampires-zombies battles. If zen 5 is about the same as 7800x3d I'm positive that's a good sign.
 
Might be out of the blue, but any expected dates/rumors on EPYC Turin so I can finally buy a used Milan-X from eBay?
They are available now on ebay. Prices are not bad, since Zen 4 server chips are so good. I would recommend them over Milan-x.
 
Back
Top