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

[Kedas]

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"

 

[biostud]

When AMD presented the 5800X3D watchdogs legion showed improvements up to 40% compared to 5900x. So wouldn't that be a pretty good guess for best case scenario in gaming?

But as the slide also showed, it varied greatly depending on which game was played, so an average doesn't really makes much sense, as you would make a decision based on which games you played.

 

[yuri69]

A single number can't represent reality w/o context.

Ideally you would benchmark every single application written for x86 and then calculate the average gain. In reality you simply want to include as many app benchmarks as possible.

 

[Mopetar]

A single number can't represent reality w/o context.

Ideally you would benchmark every single application written for x86 and then calculate the average gain. In reality you simply want to include as many app benchmarks as possible.

You don't need to measure all of them. You just need a good enough sample to get a result that's close enough to the actual average in order to be useful.

An overall average, even if it's perfect because every single program was included, isn't necessarily useful either. If it's only being used for gaming, then that's the average we care about. If games won't be used at all, but instead a

 

[biostud]

A single number can't represent reality w/o context.

Ideally you would benchmark every single application written for x86 and then calculate the average gain. In reality you simply want to include as many app benchmarks as possible.

Depends on what part of reality you are looking for. If best case scenarios are up to 40% and if newer more complex games benefits the most, then maybe it is not that important how much it improves cs:go.

Average is fine if you intend to play a wide variety of games, but personally I would be much more interested in the few games that actually showed a CPU bottleneck @ 1440p or 4k, and how many are that?

 

[moinmoin]

A single number can't represent reality w/o context.

Aside 42 anyway.

 

[MadRat]

AMD's 5800X3D strategy reminds me a bit like K6III/450 and K6III+ mobile chips that had pin-compatibility to Socket 7. Great upgrades for Socket 7 boards IF you could find a chip. Unfortunately better stuff like Slot A Athlons were already readily available by then. I fear the next BIG release is just around the corner.

 

[igor_kavinski]

I'm fatigued reading about how great 5800X3D is gonna be. AMD, get a few review samples out already!

 

[DrMrLordX]

I'm fatigued reading about how great 5800X3D is gonna be. AMD, get a few review samples out already!

Probably won't be many until April.

 

[eek2121]

10 Games, two slides. Same 15% increase.


So far these are the lists of games with 15%+ performance improvements(3D-V Cache) over a stock 5900X and 12900KS Alder Lake. Pay attention to the Tie/Even games where it does the same with the 12900K so it must be a frame cap or something. Also CS:GO Three of the CPUS ties, the 5900X, 12900K and 5800X3D so the issue is on the game.

DOTA 2: 18% Over 5900X - No info on 12900K
Monster Hunter World: 25% over 5900X - No info on 12900K
League of Legends: 4% Over 5900X - No info on 12900K
Fortnite: 17% Over 5900X - No info on 12900K
Final Fantasy XIV: 20% Over 5900X - 20% Over 12900K
Shadow of the Tomb Raider: 10% Over 5900X - 10% Over 12900K
Far Cry 6: 20% Over 5900X - 11% Over 12900K
Gears V: 12% Over 5900X - Tie/Even with 12900K
Watch Dogs Legion: 40% Over 5900X - Ties/Even with 12900K
CS:GO: Tie/Even with 5900X - Tie/Even with 12900K


View attachment 56352

View attachment 56359

You are tying to paint a bad light on 3D V Cache. It's a game changer and a generational leap. To say otherwise is just not accurate

The world is so much bigger than gaming, and gaming is so much bigger than those titles.

 

[Saylick]

Looks like the B2 stepping has better characteristics after all:

https://twitter.com/x/status/1485666028352905217

The Ryzen 5000 series of stepping B2 has no changes in function and performance. Don't expect it to have a higher frequency.

But is there no difference at all with the B2 stepping Ryzen processor? This is not necessarily true. Some big Vs such as overclocking and playing Shamino have already got the Ryzen 9 5900X processor with B2 stepping in advance and have done overclocking tests.

He did an overclocking test on ASUS ROG Crosshair VIII Extreme, and the 12-core core frequency under water cooling was increased to 5150MHz. This result is not an exaggeration, and there is no frequency change compared to the previous B0 step.


But Shamino also mentioned that the liquid cooling temperature of the B2 stepped Ryzen 9 5900X is reduced by 9 degrees after overclocking, the power consumption is reduced by 30W, and the frequency of the Prime95 multi-core grill is increased by 60MHz.

In addition, he will say in reply to netizens' questions that the memory frequency of the B2 Ryzen 9 5900X can also reach DDR4-4100, and the memory frequency seems to be better than what netizens mentioned.

 

[Mopetar]

The world is so much bigger than gaming, and gaming is so much bigger than those titles.

This is a thread about v-cache performance in games. We already know that it won't benefit every workload, but if you have one that doesn't benefit then don't buy this CPU.

As long as the games used to calculate that average are a good representative sample of the total set of games, the average will be close enough to correct for useful purposes, just like a poll of 500 people will give you a close enough result assuming the people polled aren't disproportionately selected.

 

[Zucker2k]

You know, the more I think about the 5800X3D, the more I think "what a great chip this will make for the consoles." Consoles will get a huge boost with this tech.

 

[eek2121]

You know, the more I think about the 5800X3D, the more I think "what a great chip this will make for the consoles." Consoles will get a huge boost with this tech.

LOL that will never happen. This chip is likely to cost more than an entire console.

 

[lightmanek]

LOL that will never happen. This chip is likely to cost more than an entire console.

Never say never
We will get there and beyond with 3nm.

 

[Mopetar]

You know, the more I think about the 5800X3D, the more I think "what a great chip this will make for the consoles." Consoles will get a huge boost with this tech.

Honestly if AMD could figure out how to make a unified SLC for their APUs it would likely be good boost for consoles to add some stacked cache. Even though it wouldn't turn them into top-tier PC equivalents, it probably would help with minimum framerates and eliminating some of the hiccups in gameplay.

 

[RnR_au]

Looks like the B2 stepping has better characteristics after all:

In the twitter thread, 1usmus is claiming 5250Mhz...

https://twitter.com/x/status/1485908205121191936

 

[Tarkin77]

new press samples? why?

 

[Thibsie]

new press samples? why?

Cherry picking purposes ?

 

[leoneazzurro]

Maybe it's a sample of a 5800X3D in the case of 1usmus, the other test did not mention press samples...

 

[moinmoin]

new press samples? why?

Samples for everyone! (Except us.)

 

[Mopetar]

new press samples? why?

Might also be that AMD is going to do some kind of refresh. They've done this in the past with XT models that had slight clock bumps.

 

[igor_kavinski]

Maybe it's a sample of a 5800X3D in the case of 1usmus, the other test did not mention press samples...

AMD could be monitoring these forums and they heard me

AMD, please send a sample my way too. I can't promise I will be gentle with it, but I will sure give it all my love and attention.

 

[Det0x]

In the twitter thread, 1usmus is claiming 5250Mhz...

https://twitter.com/x/status/1485908205121191936

Pretty sure 1usmus is talking about his first B0 sample.. (which was seeded to press before zen3 release)

Anyway both my B0 2040 5950x and 2130 5950x samples can do ST 5250 on custom watercooling so there is no news here.

Everything after 2143 is B2 samples which have been sold for months already.. And sadly it seems like they are no better than B0 samples.

 

[coercitiv]

Good to know for people reading more stories about improved parameters in B2 stepping.

https://twitter.com/x/status/1486646970361909248

There's this story going around that Vermeer B2 uses 30W less power and has 9C lower temps. Incidentally those are the same numbers I reported in my 5900 B2 OC guide. Except that the -30W and -9C was from using Shamino's Voltage Suspension feature on C8E ... not from B2 stepping

 

[moinmoin]

So changes in B2 are essentially coming from a tweaked firmware?