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]

So, in theory, the games where the 5800X3D has the big wins would also be games where ADL would make significant gains with DDR5? That makes sense to me. I don't think whatever price differences between the 2 combos you mentioned would be a huge factor for anyone looking at these for the absolute best in gaming, since they'll probably be rocking a 3090 or 3090 Ti as well, so a $50 price difference is literally loose change at the bleeding edge of PC gaming systems which costs thousands of dollars.

Based on this review of the 12900KS showing gains up to 20% with DDR5, I'll wait for comparisons with DDR5 based ADL systems before drawing my conclusions, though it seems with DDR4 a 5800X3D at worst ties with a 12900K and beats it handily in a few games which is a great result, though in line with my expectations all along.

I just checked Techspot and Techpowerup for their memory scaling articles but neither tested FFXV. Would like to see that game tested to see if that too is memory heavy.

 

[Heartbreaker]

Early gaming review of 5800X3d:

AMD Ryzen 7 5800X3D Review – The last gaming gift for AM4

AMD's last gaming gift for AM4 platform. Ryzen 7 5800X3D review. Benchmarks and work in progress of the new AMD's gaming focused product.

xanxogaming.com

It's vs a 12900K, with both on DDR4-3200.

Very close on most games, with big wins on Witcher 3, SotTR, FFXV for 3D cache.

Though you kind of need to be running a 3080 or better for this to matter, and my end of the market is more 12400 vs 5600, paired with something like an RX 6600.

 

[MadRat]

Wow, like an old V8 with a nice cam versus a supercharged V4. Sure they both have similar power at their peak but the V8 was made to sustain peak work for longer. Those average FPS scores are very eye-popping. I would have expected it to be closer. Makes me think Zen 4 is going to shatter the 5800x3D performance once it hits stride.

 

[nicalandia]

Alright, dig in!

AMD Ryzen 7 5800X3D Review – The last gaming gift for AM4

AMD's last gaming gift for AM4 platform. Ryzen 7 5800X3D review. Benchmarks and work in progress of the new AMD's gaming focused product.

xanxogaming.com

You beat me to it.. Thanks..! ChanCho Gaming milked that as much as they could... Such Cocktease..!

 

[Timorous]

Early gaming review of 5800X3d:

AMD Ryzen 7 5800X3D Review – The last gaming gift for AM4

AMD's last gaming gift for AM4 platform. Ryzen 7 5800X3D review. Benchmarks and work in progress of the new AMD's gaming focused product.

xanxogaming.com

It's vs a 12900K, with both on DDR4-3200.

Very close on most games, with big wins on Witcher 3, SotTR, FFXV for 3D cache.

Though you kind of need to be running a 3080 or better for this to matter, and my end of the market is more 12400 vs 5600, paired with something like an RX 6600.

Metro Exodus and Control also show a 15% and 10% increase in 1% lows respectively at 1080p which is pretty good.

Death Stranding has a 240 FPS cap as well so there the AMD system is very very close to running at the limit in 1080p with a 234 FPS average.

 

[tamz_msc]

12900K with tuned DDR5 memory will be faster than 5800X3D with tuned DDR4. Except in edge cases like StarCraft II which is heavily bound by memory latency. Modern open world and RTS games that rely on streaming assets from memory will be faster on Alder Lake + tuned DDR5.

i2Hard's DDR4 vs DDR5 memory scaling benchmarks seem to indicate this.

 

[jpiniero]

12900K with tuned DDR5 memory will be faster than 5800X3D with tuned DDR4. Except in edge cases like StarCraft II which is heavily bound by memory latency. Modern open world and RTS games that rely on streaming assets from memory will be faster on Alder Lake + tuned DDR5.

i2Hard's DDR4 vs DDR5 memory scaling benchmarks seem to indicate this.

I think it's more like that older games with very high frame rates will benefit most. There's going to be titles where the 5800X3D is faster than Zen 4 even.

 

[tamz_msc]

I think it's more like that older games with very high frame rates will benefit most. There's going to be titles where the 5800X3D is faster than Zen 4 even.

Nah, old games like CS:GO already benefit from 32 MB unified cache, slapping an additional 64 MB will not improve performance, except perhaps StarCraft II. AMD's own benchmarks show that the difference in CS:GO is negligible.

 

[nicalandia]

12900K with tuned DDR5 memory will be faster than 5800X3D with tuned DDR4.

If you believe that 12900K Was Handicapped by the DDR4 and AMD was not. Then you just don't know how AMD Zen 3 benefit from Tunned RAMs

 

[tamz_msc]

If you believe that 12900K Was Handicapped by the DDR4 and AMD was not. Then you just don't know how AMD Zen 3 benefit from Tunned RAMs

Not to the extent that Alder Lake does with tuned DDR5:



In a game like CP2077, Alder Lake is over 50% faster than Zen 3 with tuned DDR4. I don't see 5800X3D overcoming this.

Screenshot from here:

 

[Heartbreaker]

If you believe that 12900K Was Handicapped by the DDR4 and AMD was not. Then you just don't know how AMD Zen 3 benefit from Tunned RAMs

But the larger 3D cache better insulates AMD from slower RAM, that's kind of the whole point.

 

[nicalandia]

Not to the extent that Alder Lake does with tuned DDR5:

Zen3D with Tune RAM will have no issue overcoming that. In any event, the 12900K/KS will just look really bad when compared to 5800X3D in any gaming metric(Price/Performance, Power/Performance) if you have to shell another $400 for Good DDR5 RAM.

 

[Timorous]

Not to the extent that Alder Lake does with tuned DDR5:

View attachment 59892

In a game like CP2077, Alder Lake is over 50% faster than Zen 3 with tuned DDR4. I don't see 5800X3D overcoming this.

Screenshot from here:

TPU get different results.

720P:

1080P:

Not so sure I would trust random YouTubers.

 

[tamz_msc]

TPU get different results.

Not so sure I would trust random YouTubers.

I urge you to take a look at the video. TPU looks amateurish compared to them.

 

[exquisitechar]

12900K with tuned DDR5 memory will be faster than 5800X3D with tuned DDR4. Except in edge cases like StarCraft II which is heavily bound by memory latency. Modern open world and RTS games that rely on streaming assets from memory will be faster on Alder Lake + tuned DDR5.

i2Hard's DDR4 vs DDR5 memory scaling benchmarks seem to indicate this.

I agree, the 5800X3D's gaming performance is not particularly earth shattering and 12900K with tuned DDR5 will be faster, it seems. Although DDR5 is very expensive, so being very fast with relatively cheap DDR4, out of the box, is pretty good for the 5800X3D, I suppose.

 

[nicalandia]

I urge you to take a look at the video. TPU looks amateurish compared to them.

Is Someone grasping at straws?

 

[tamz_msc]

Zen3D with Tune RAM will have no issue overcoming that. In any event, the 12900K/KS will just look really bad when compared to 5800X3D in any gaming metric(Price/Performance, Power/Performance) if you have to shell another $400 for Good DDR5 RAM.

That doesn't have any bearing on the question 'which is the fastest processor for gaming?'*

*when you account for the platform-specific tuning that is available.

 

[tamz_msc]

Is Someone grasping at straws?

The ones grasping at straws are those who ignore valid data.

 

[tamz_msc]

I agree, the 5800X3D's gaming performance is not particularly earth shattering and 12900K with tuned DDR5 will be faster, it seems. Although DDR5 is very expensive, so being very fast with relatively cheap DDR4, out of the box, is pretty good for the 5800X3D, I suppose.

Yeah, as a swansong for AM4, 5800X3D is a great CPU for what it does.

 

[nicalandia]

The ones grasping at straws are those who ignore valid data.

Grasping at Straws does not suit you at all.

Just say the generic "Let's wait for Valid Benchmarks" while discrediting TPU at the same time.

 

[tamz_msc]

Grasping at Straws does not suit you at all.

Just say the generic "Let's wait for Valid Benchmarks" while discrediting TPU at the same time.

TPU CP2077 1080p data is GPU-bound. It's worthless. 11 FPS difference going from 6700K to 10900K. That's all it should take to judge the quality of their data.

 

[Hans Gruber]

I agree, the 5800X3D's gaming performance is not particularly earth shattering and 12900K with tuned DDR5 will be faster, it seems. Although DDR5 is very expensive, so being very fast with relatively cheap DDR4, out of the box, is pretty good for the 5800X3D, I suppose.

It's kind of why I said that any Zen 3 CPU should be tested against Alder Lake CPU's with minimum 3600mhz Cas 14 to 3800mhz Cas 16. DDR5 is better but high clocked memory can still hold it's own vs. the new DDR5 kits. Within a year, DDR5 will leave DDR4 in the dust. Remember the 5800x3D is a demonstrator CPU for 3D v-cache. A nice to have feature for gaming that will probably be standard issue in Zen 4 CPU's along with onboard integrated GPU's from AMD.

The addition to onboard graphics in Zen4 will be a game changer for AMD. It's something few need but comes in handy when in between GPU's. 3d v-cache and onboard graphics are features that add value to a CPU.

 

[carrotmania]

In a game like CP2077, Alder Lake is over 50% faster than Zen 3 with tuned DDR4. I don't see 5800X3D overcoming this.

You didnt look at that video and feel something was "off"? I havent seen any other review where zen3 is 50% slower than ADL, tweaked or not.

etc etc

 

[Timorous]

I urge you to take a look at the video. TPU looks amateurish compared to them.

TPU says something you don't like is more accurate.

Techspot also did some scaling testing at 1080p Medium.

And from there launch review we see when using 1080p High.

The 5800X from the 12700K review scored 133 average and 106 min.

Seems like i2Hard is the outlier.

 

[tamz_msc]

You didnt look at that video and feel something was "off"? I havent seen any other review where zen3 is 50% slower than ADL, tweaked or not.

etc etc

Why are people linking GPU-bound results in an argument about CPU gaming performance?