Info 64MB V-Cache on 5XXX Zen3 Average +15% in Games

Page 123 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Kedas

Senior member
Dec 6, 2018
355
339
136
Well we know now how they will bridge the long wait to Zen4 on AM5 Q4 2022.
Production start for V-cache is end this year so too early for Zen4 so this is certainly coming to AM4.
+15% Lisa said is "like an entire architectural generation"
 
Last edited:
Reactions: Tlh97 and Gideon

moinmoin

Diamond Member
Jun 1, 2017
4,993
7,763
136
I think I read that Zen 3 actually only gets the minimum latency out of the L3 with a set size of 8 MB or less since it overflows the TLB at that size. If they enhance the TLB (very likely), then average L3 latency could go down significantly, even with no L3 size increase.
Great point. This is what the TLB penalty looks like in AT's Zen 3 deep dive (by Andrei Frumusanu):



A single threaded tech demo in 2022? That seems unlikely to be the future of anything.
There's actually no Matrix demo for PC (yet?). The only thing that exists is project data of the city than can be compiled (which obviously wouldn't be optimized beyond anything UE5 offers at stock).
 

Hitman928

Diamond Member
Apr 15, 2012
5,567
8,717
136
Those numbers do look weird but it also also be Single CCD vs dual CCD at least with the Zen 3 parts.

There are sometimes when the 5600X and 5800X will out perform the 5900X and 5950X due to the latency in single vs Dual CCD aslong as your workload is not bottlenecked by core count.

The CCX/CCD penalty in a 2700x is way worse than on a 5900X. There really shouldn't be any workload where a 2700x outperforms a 5900x and this holds especially true for gaming. There is a major issue going on on the AMD CPU side. It looks like have multiple CCDs is making it worse, but there is a bigger underlying issue here.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,736
14,767
136
The CCX/CCD penalty in a 2700x is way worse than on a 5900X. There really shouldn't be any workload where a 2700x outperforms a 5900x and this holds especially true for gaming. There is a major issue going on on the AMD CPU side. It looks like have multiple CCDs is making it worse, but there is a bigger underlying issue here.
How do you trust a picture of a Russian (??) benchmark ? Its not a link to somewhere on the web, so it may not be real ?? I am not accusing, just saying what if .
 
Reactions: Tlh97 and Drazick

jpiniero

Lifer
Oct 1, 2010
14,823
5,440
136
There's actually no Matrix demo for PC (yet?). The only thing that exists is project data of the city than can be compiled (which obviously wouldn't be optimized beyond anything UE5 offers at stock).

The CitySample part is the same thing as what you get on the consoles. You just don't get the chase scene.

How do you trust a picture of a Russian (??) benchmark ? Its not a link to somewhere on the web, so it may not be real ?? I am not accusing, just saying what if .


They obviously didn't optimize it that much. It's just meant to be an example for developers.
 
Reactions: Tlh97 and Markfw

Makaveli

Diamond Member
Feb 8, 2002
4,755
1,145
136
The CCX/CCD penalty in a 2700x is way worse than on a 5900X. There really shouldn't be any workload where a 2700x outperforms a 5900x and this holds especially true for gaming. There is a major issue going on on the AMD CPU side. It looks like have multiple CCDs is making it worse, but there is a bigger underlying issue here.

That is why I specifically said in the Zen 3 line and didn't reference the 2700X. And this is still an optimized PC tech demo so I wouldn't take those results as gospel.
 
Last edited:

Hitman928

Diamond Member
Apr 15, 2012
5,567
8,717
136
That is why I specifically said in the Zen 3 line and didn't reference the 2700X. And this is still an optimized PC tech demo so I wouldn't take those results as gospel.

I think you mean un-optimized, which I agree with. I would say it's more than just not optimized though and needs a bug fix pass for AMD CPUs. Either way, I think we agree that this shouldn't be used as a measuring stick for CPU performance which is what the poster I originally responded to was doing.
 
Reactions: Tlh97 and Makaveli

nicalandia

Diamond Member
Jan 10, 2019
3,331
5,282
136
So due to Covid the Anand Tech Review of 5800X3D was postponed.

"Unfortunately we're not going to have a Ryzen 7 5800X3D review out for today. Despite herculean efforts from @gavbon86 to get the necessary testing done in time, COVID has won this round. We'll have something next week."

 
Last edited:

Saylick

Diamond Member
Sep 10, 2012
3,372
7,103
136
So due to Covid the Anand Tech Review of 5800X3D was postponed.

"Unfortunately we're not going to have a Ryzen 7 5800X3D review out for today. Despite herculean efforts from @gavbon86 to get the necessary testing done in time, COVID has won this round. We'll have something next week."

I'm pleasantly surprised there even is a forthcoming review... Anandtech isn't what it used to be these days.
 

nicalandia

Diamond Member
Jan 10, 2019
3,331
5,282
136
Last edited:

nicalandia

Diamond Member
Jan 10, 2019
3,331
5,282
136
My Guess from the Milan-X Delid is that they were expecting three separate structures on top of the CCD die, but only found and single structure or what it looks like one since it's not a IR or other wave length camera.

According to the source they believe AMD is placing another silicon support on top of the 3D V-Cache/Support structures. This to me seem rather unlikely due to the Hight issues, the CCD does not seem to be any thicker than a regular CCD(AMD has said that both the L3 die and structures are thinned down) so what I think it's happening here is that after both the L3$ die and Support are placed on top of the CCDs they are sand down making it appear to be a single support instead of being three structures.







Edit.

Never Mind. It is a Single Support Silicon on top of a Thinned Down CCD

 
Last edited:

nicalandia

Diamond Member
Jan 10, 2019
3,331
5,282
136
Plenty of good reviews already.
I can't wait for AnandTech SPEC06/2017 numbers that are not done at ISO speed.

The ONLY time AnandTech Review team did a test with SPEC at ISO Speed(as to compare IPC) was when they were testing Skylake 14nm with Skylake on 10nm... The rest of the time they run SPEC at stock speeds
 
Last edited:
Jul 27, 2020
17,712
11,499
106
What was the increase in L3 cycles over the regular 5800X? 3 some cycles?
4, if Semiaccurate is to be believed.


On the latency front the L3 cache on the -3D is about 4 cycles slower than the non-3D parts, something AMD went to very tortured lengths to try and minimize in the briefings.

He seems annoyed that AMD released this as a product.
 

dr1337

Senior member
May 25, 2020
384
635
136
4, if Semiaccurate is to be believed.




He seems annoyed that AMD released this as a product.
hmm he seems to write very much like the person behind userbenchmark

"The 3D V-Cache is a really good idea for number of server markets but the flaws mean it has no real market on the consumer side. Sure it can put up a few good numbers but the 5800X3D is so ill-suited for the task at had it makes us question why AMD bothered. Oh yeah marketing. "-SA article above

"AMD’s marketers continue to show more interest in this year’s bonuses than the longevity of the brand. Instead of focussing on real-world performance, they attempt to dupe consumers with benchmark busting headlines. " -UB 5800x3d bio

Maybe its more of a niche chip than I realize but I don't see where this sentiment comes from.
 
Reactions: lightmanek
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/    |