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"

 

[Timorous]

Can they still keep saying that it's the "fastest" at this point?

Depends how it does in lots of other games. Might very well average out as fastest in a 20 or 30 game test suite.

 

[epsilon84]

Depends how it does in lots of other games. Might very well average out as fastest in a 20 or 30 game test suite.

Yeah, I'll give AMD the benefit of the doubt here. I'm sure they have a 12900K/KS system in house they can compare against. Would be silly to make such a claim and have reviewers come to a different conclusion

 

[Makaveli]

Tbh, I wouldn't put too much faith in those results. Brand new channel, literally less than 24 hours old, no subs lol.

🙂 i'm still waiting for the official reviews tomorrow this is just adding more data to the thread.

 

[Markfw]

Can they still keep saying that it's the "fastest" at this point?

Most of the "pre-benchmarks" have it winning everything. A couple of outliers , maybe have the 12900ks winning a couple.. Tomorrow it looks like all the sites will come out with benchmarks, then we will really know.

 

[nicalandia]

Can they still keep saying that it's the "fastest" at this point?

It is the World's Best/Fastest GAMING CPU. There is no other way around that

 

[BTRY B 529th FA BN]

🙂 i'm still waiting for the official reviews tomorrow this is just adding more data to the thread.

There's an NDA lift on the 14th?

 

[lobz]

Yet another ignorant take.

Not really, I'm just still tracing them rays.

Especially when you 'accidentally' 'missed' my take on the CPU, which literally goes: 'would have been nice of them to have some nice DDR5 just to show the 12900K in its best light too, but you can't really argue about the cost efficiency - oh, how the turntables (sic!) - of the 5800X3D here'

Man, isn't this forum an exhaustingly ignorant place?

 

[jpiniero]

I want to see how well it does in the Matrix demo.

 

[lightmanek]

I want to see how well it does in Quake 2 software renderer ...
(fully expecting to be unimpressive due to engine optimization made for CPUs with 16/32Kb of cache).

 

[Borealis7]

i've read the TPU review, and i just dont see the point. IMO, this is a CPU without a target demographic.
Nobody plays games at 720p, at higher resolutions i can't justify the 100$ price increase for the performance gains, especially not at 4K where the CPU matters much less.
don't get me wrong, its an impressive technical achievement with the specific gains and power draw, but its not a revolution. more like an AMD "proof of concept" CPU.

 

[Grabo]

i've read the TPU review, and i just dont see the point. IMO, this is a CPU without a target demographic.
Nobody plays games at 720p, at higher resolutions i can't justify the 100$ price increase for the performance gains, especially not at 4K where the CPU matters much less.
don't get me wrong, its an impressive technical achievement with the specific gains and power draw, but its not a revolution. more like an AMD "proof of concept" CPU.

I agree that it seems like a proof of concept and that it doesn't appear to give any advantage at 4K (which was somewhat expected as that resolution is traditionally a vertical wall in graphs across different cpus) but for 1440p there are gains in some games; all in all this seems like the best for-games CPU for AM4. Depending on what you already have, which is a wide parameter since AM4 has been around for a while, and whether you game a lot, this CPU might be better than replacing your mobo, cpu and RAM(RAM because an oft-repeated point about socket 1700 is that it isn't a dead end because raptor lake will reuse it, well I would get a motherboard with DDR5).

A better graphics card is often a better deal for games but the local market isn't like the U.S market when it come to gpu prices.

 

[lobz]

i've read the TPU review, and i just dont see the point. IMO, this is a CPU without a target demographic.
Nobody plays games at 720p, at higher resolutions i can't justify the 100$ price increase for the performance gains, especially not at 4K where the CPU matters much less.
don't get me wrong, its an impressive technical achievement with the specific gains and power draw, but its not a revolution. more like an AMD "proof of concept" CPU.

You realize your statement could be applied for every single CPU over $250?

 

[gdansk]

i've read the TPU review, and i just dont see the point. IMO, this is a CPU without a target demographic.

It seems to be good at gaming. It's cheaper than 12900K. That's a target market.
It is an upgrade for existing AM4 users. There's another.
But with new PCs, you're better off buying a midrange CPU and putting the saved $200-400 toward the GPU. And since there's no budget Zen3D that still means Alder Lake.

 

[Karnak]

Nobody plays games at 720p

That's not the point using 720p to test CPUs. It's to isolate the impact of a GPU which you dont want since you want to know of what the CPU itself is capable of.

Resolution doesnt matter for the CPU. If CPU A get's 90fps in 720p, then CPU A get's the exact same 90fps in 4K. If you're GPU is limiting at 70fps though you wont know that.

And I think it's safe to say that a CPU will last a lot longer than a GPU in most cases. So it's good to know if you're upgrading your GPU in the future what you still can expect from your CPU. For that you need testing in lower res.

Although you could just bench games like Total War, Cities: Skylines or Anno and even get CPU bottlenecked in 4K cause those games want every single fps from your CPU that they can get and even in higher res the GPU just doesn't matter at some point. Don't know why reviews are still sticking to this 3xxfps games nonsense or heck even CS:GO with over 7xxfps. Who literally cares about that if there are games out there where you get like sub 60fps today cause all CPUs can't deliver more (while the GPUs on the other side can).

 

[Timorous]

i've read the TPU review, and i just dont see the point. IMO, this is a CPU without a target demographic.
Nobody plays games at 720p, at higher resolutions i can't justify the 100$ price increase for the performance gains, especially not at 4K where the CPU matters much less.
don't get me wrong, its an impressive technical achievement with the specific gains and power draw, but its not a revolution. more like an AMD "proof of concept" CPU.

Nobody tests the kinds of games where 4K is actually easy for the GPU but the simulation itself is held back by the CPU. Late game Cities Skylines with a high pop city. Late game Stellaris. 4X Turn times (outside of Civ 6 which on occasion gets a test) and so on. Lots of MMO games are also very CPU limited but they are really hard to test reliably due to patches and the nature of online games so they also don't get tested.

Currently you might get the Factorio bench and you might get Civ 6 Turn time but that is about it for non FPS gaming metrics. Outside of that nada and often the games that do get tested are not CPU heavy so it is kinda pointless.

This 'Only the GPU matters at 4K' nonsense does need to die though because it is only true for a subset of games, not all games.

 

[JoeRambo]

Nobody tests the kinds of games where 4K is actually easy for the GPU but the simulation itself is held back by the CPU. Late game Cities Skylines with a high pop city. Late game Stellaris. 4X Turn times (outside of Civ 6 which on occasion gets a test) and so on. Lots of MMO games are also very CPU limited but they are really hard to test reliably due to patches and the nature of online games so they also don't get tested.

Currently you might get the Factorio bench and you might get Civ 6 Turn time but that is about it for non FPS gaming metrics. Outside of that nada and often the games that do get tested are not CPU heavy so it is kinda pointless.

I have been harping about this since original Zen came out. Even in "average" FPS department - frame rate stability and FPS minimums and percentiles are also (if not more) important.
The reason: it is easier for reviewers to automate the FPS tests and have them fully reproducible and with little variance between the runs. While testing something like sim speeds in Stellaris or Anno 1800 or those turn based game turn times is hard to setup and even harder to reliably reproduce.

 

[DrMrLordX]

i've read the TPU review, and i just dont see the point. IMO, this is a CPU without a target demographic.

The target demographic:

1). Mostly just plays games
2). Wants something fast at any resolution
3). Doesn't want to spend ~$700 + DDR5-6000 tax
4). May already have an AM4 motherboard and a 6c or 8c Zen2/Zen3 CPU

 

[biostud]

A faster CPU will last longer than than a slower, and whether it is worth the extra cost is up to you. Benchmarking as many games as possible at low resolution will give a good indication on how different game engines perform with a CPU, benchmarking at resolutions you actually play, will show you that any modern CPU from $250 and up will give you the same fps, as your video card will be be the bottleneck. Then why buy a more expensive CPU? because you also do other stuff on your computer, and the extra CPU power might matter next time you buy a new video card (or it might not 😛)

But pretty crazy that you can get the 5950X for $549 @ Amazon.

 

[BigDH01]

Nobody tests the kinds of games where 4K is actually easy for the GPU but the simulation itself is held back by the CPU. Late game Cities Skylines with a high pop city. Late game Stellaris. 4X Turn times (outside of Civ 6 which on occasion gets a test) and so on. Lots of MMO games are also very CPU limited but they are really hard to test reliably due to patches and the nature of online games so they also don't get tested.

Currently you might get the Factorio bench and you might get Civ 6 Turn time but that is about it for non FPS gaming metrics. Outside of that nada and often the games that do get tested are not CPU heavy so it is kinda pointless.

This 'Only the GPU matters at 4K' nonsense does need to die though because it is only true for a subset of games, not all games.

Agreed. I really don't care about how these CPUs do in FPS titles because I want to play at 4K and the GPU is going to be the bottleneck anyway. I want to know how it does late game Hearts of Iron 4 where the CPU might actually matter.

 

[MadRat]

The important question is can it hit 16,000 fps in the original Doom?

 

[Asterox]

BorderLands 3 1440p, that FPS jump is Klingon crazy.

R7 5800X, 95fps
R7 5800X3D, 136fps

 

[Insert_Nickname]

The important question is can it hit 16,000 fps in the original Doom?

Wouldn't 16384 fps be more appropriate in this particular context? 😉

 

[BTRY B 529th FA BN]

BorderLands 3 1440p, that FPS jump is Klingon crazy.

"What did you say, Hueman"

 

[nicalandia]

Nobody plays games at 720p, at higher resolutions i can't justify the 100$ price increase for the performance gains,

I can't understand some people here. Yeah Nobody games at 720P(not the point of the benchmark by the way), And whoever plays at 4K is not bottle necked by the CPU but by the GPU, pair any decent CPU of the last 4 years with a 3090 Ti and be happy..

But this CPU, It's the Gaming King at 1440P where most people would agree that is a very impressive piece of technology and from available data, It beats the 12900KS at 1080P and 1440P(TPU and Chancho Gaming)

 

[epsilon84]

I'm curious if anyone here is actually planning to get a 5800X3D? I'm unsure who would be the most suitable target for this kind of chip. Yes, I realise its targeted at high end gamers, but high end gamers generally already own a Zen 3 or ADL CPU and I don't think the 5800X3D is *that* much faster to make the switch worthwhile, especially for those running ADL systems. For AMD, those already running 5900X/5950Xs may be reluctant to give up cores for higher gaming IPC. Is anyone here actually planning to move from a 5900X/5950X to a 5800X3D?

Of course, we also have the older gen AM4 gamers, those not currently running Zen 3. Generally speaking, until Zen 3 arrived, AM4 was the budget gamers choice, with chips like the 1600/2600/3600 providing great value at the $150-$200 pricepoints, despite not directly outperforming the Intel CPUs from its era in gaming. I'm not sure how many of those budget conscious gamers will now be willing to spend $450, basically 2-3 times the outlay of their original Zen 1/2 CPUs, when a $200 Ryzen 5600 gets within ~15% of the gaming performance of a 5800X3D and realistically, at higher res or with more mainstream GPUs, the gap closes further.