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

Page 710 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

tsamolotoff

Member
May 19, 2019
184
319
136
Thank god, this will hopefully end the thousands of "my CPU is running too hot" posts in German forums that I had to endure for Ryzen 7000 during the last two years.
It'd matter if it provided tangible benefit in performance, but most people will just open their faviourite tech-youtuber (GN or HU), see their thumbnails and won't buy the CPUs. AMD simply shot into both their legs and the dangling little thing between them as well...
 

Asterox

Golden Member
May 15, 2012
1,039
1,823
136
Where Zen 5 actually shines is temperature:
View attachment 104662
That's -35°C from 7700X to 9700X and still -15°C when both are running at 142W PPT. Thank god, this will hopefully end the thousands of "my CPU is running too hot" posts in German forums that I had to endure for Ryzen 7000 during the last two years.
Crazy lower numbers, from LTT Youtube test.

2024-08-07_171251.jpg

 

inquiss

Senior member
Oct 13, 2010
201
276
136
AVX-512 is only a small part of RPCS3. A big uplift on it won't translate to the same uplift for the PS3 emulator

? AMD is fine. Core will do well on servers, it will be competitive on mobile and they'll have a cost advantage there. And it's a much more efficient core for Office/General DT tasks. The only public that will be disappointed will be the gaming one.

But that doesn't mean they'll need to rush Z6. If anything, they need to make sure Z6 improves radically over Z5.
And these are not the gaming chips. That's the cache ones. Gamers would be picking the cache code anyway
 

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
Even without this no one really uses or benchmarks the CPUs with JEDEC memory, it tanks AM5 performance disproportionately hard and isn't really used in real life scenario (unless you build home server with some sort of cheap ECC ram)
5600Mt/s is the official spec for those CPUs, so it make sense to make comparisons with the guaranted RAM speed.

Beside that s not the point, the guy selected the games such that CB tests results are the same as PCGH whose reviewer is known to trick the numbers in Intel s favour, FI in his 7950X review he doctored the tests such that the 14900K is 15% faster in MT while at Computerbase it barely matched the 7950X even when using 330W.
 

Hitman928

Diamond Member
Apr 15, 2012
6,187
10,694
136
5600Mt/s is the official spec for those CPUs, so it make sense to make comparisons with the guaranted RAM speed.

Beside that s not the point, the guy selected the games such that CB tests results are the same as PCGH whose reviewer is known to trick the numbers in Intel s favour, FI in his 7950X review he doctored the tests such that the 14900K is 15% faster in MT while at Computerbase it barely matched the 7950X even when using 330W.

For Intel, official spec is actually ddr5 4400 for any boards with 4 slots, even if only two of the slots are populated. That’s something that came out of the recent instability saga for Intel. I highly doubt the people saying only official memory speeds should be tested would actually want to follow that spec though.
 

Tup3x

Golden Member
Dec 31, 2016
1,086
1,085
136
So, if this is the result for non-X3D version, I somewhat doubt that X3D variant would be much better. At least it wouldn't be worth it for Zen 4 users.

In any case... Two years and this is the result?
 

soresu

Diamond Member
Dec 19, 2014
3,230
2,515
136
Bulldozer was a departure from Intel-like cores and it suffered on Windows way more than on Linux
IIRC it took a while for the Windows scheduler to properly support the Bulldozer CMT architecture.

MS have a tendency to gatekeep such things within major OS updates, while with Linux it just happens whenever the code is ready.
 
  • Like
Reactions: igor_kavinski

biostud

Lifer
Feb 27, 2003
18,700
5,434
136
So, if this is the result for non-X3D version, I somewhat doubt that X3D variant would be much better. At least it wouldn't be worth it for Zen 4 users.

In any case... Two years and this is the result?
As already mentioned, only if the X3D versions are not downclocked like the 7800X3D, which might be a possibility with the low TDP of the 9700X.
 
  • Like
Reactions: Joe NYC

tsamolotoff

Member
May 19, 2019
184
319
136
From what I have seen even in games the V$ is very variable in benefit ranging from nothing to 20+% improvement.
Yes, those extra % are basically free in terms of power or architecture, that's what I've meant. Also, it's not 20%? it's roughly 30% in cs2 (posted my own comparisons a week ago or so in this thread), around 50%+ in tarkov and pubg, and so on. Most people simply don't know how to test CPUs and their benchmarks are basically GPU tests instead of CPU tests.
 
  • Like
Reactions: Tlh97 and Racan

soresu

Diamond Member
Dec 19, 2014
3,230
2,515
136
I think the V cache helps basically always in gaming
AFAIK the cache increase only actually helps if the game is moving data resources around significantly in flight.

Some game engines are better than others at managing data movement, so it's completely within reason that some games will benefit far less from cache increases like X3D.

Basically V cache is highlighting lazy programmers as much as anything else.

I'd be interested to see V$ benefits on day 1 games vs after a couple years of patches to refine perf.
 

IEC

Elite Member
Super Moderator
Jun 10, 2004
14,441
5,434
136
Unlike Zen 4 which benefited from DDR4-->DDR5 jump, Zen 5 is still stuck on similar DDR5 speeds. And we already knew Zen 4 was memory constrained in a lot of workloads which is why the 3D vCache variant does so well in some instances. Looks like same still applies to Zen 5 but it's likely to hit the bottleneck even harder due to mArch differences.

Running JEDEC speeds seems like it gimps performance even more vs Zen 4.

1723046468818.png
 

gdansk

Platinum Member
Feb 8, 2011
2,966
4,498
136
Bulldozer was a departure from Intel-like cores and it suffered on Windows way more than on Linux. For some reason Zen 5 seems to behave similarly.
I do not think it is operating system related. It's mainly the tested workloads.
Bulldozer was better in Linux because of its unusual scheduling requirements. That's not the case here. From the operating systems perspective these 1 CCD 1 CCX parts need no scheduling changes. It is effectively the same structure since 2020 with Zen 3.
 

Hitman928

Diamond Member
Apr 15, 2012
6,187
10,694
136
AFAIK the cache increase only actually helps if the game is moving data resources around significantly in flight.

Some game engines are better than others at managing data movement, so it's completely within reason that some games will benefit far less from cache increases like X3D.

Basically V cache is highlighting lazy programmers as much as anything else.

I'd be interested to see V$ benefits on day 1 games vs after a couple years of patches to refine perf.

I can't remember seeing any tests where the V-cache didn't provide a gaming benefit. There are tests where it wasn't enough benefit to overcome the frequency advantage of the non-Vcache parts. I'm sure there are outliers though.
 

soresu

Diamond Member
Dec 19, 2014
3,230
2,515
136
I can't remember seeing any tests where the V-cache didn't provide a gaming benefit. There are tests where it wasn't enough benefit to overcome the frequency advantage of the non-Vcache parts. I'm sure there are outliers though.
Do they not just fix clock frequency on both to get an accurate reading on the perf delta?

Also I was under the impression that while the 5800X3D clock regression issue was very significant that the 7x00X3D SKUs were much better for this?
 

Hitman928

Diamond Member
Apr 15, 2012
6,187
10,694
136
Do they not just fix clock frequency on both to get an accurate reading on the perf delta?

Also I was under the impression that while the 5800X3D clock regression issue was very significant that the 7x00X3D SKUs were much better for this?

Vast majority did not test iso clock, maybe someone did.

Frequency gap between vanilla and Vcache parts grew from Zen 3 to Zen 4.
 
Last edited:

CakeMonster

Golden Member
Nov 22, 2012
1,502
659
136
I was pretty adamant on getting a 9950X, even if the gain over 7950X was mediocre. Now I'm in a bit of a bind though, because I really dislike the dual-CCX thread prioritization mess and I still consider it very likely that a 9950X3D would have heterogenous cores.