• Happy Holidays Guest!

    We want to announce that we will be working at reduced staffing for the holidays. Specifically Monday the 23rd until Jan 2nd.

    This will affect approval queue times and responses to support tickets. Please adjust your plans accordingly and enjoy yourselves this holiday season!

  • Hi Guest!

    Please be aware that we have released a critical security patch for VaM. We strongly recommend updating to version 1.22.0.7 using the VaM_Updater found in your installation folder.

    Details about the security patch can be found here.

Benchmark Result Discussion

Hello, maybe someone knows what is better to choose for physics simulation, for example 3-4 characters in a scene with soft physics? I have a choice between 5700x and 5700x3d. I wonder what is better, a higher frequency or a larger cache.
 
Dude, just take a 5800X3D and you have the fastest you can get for AM4. No more worries.

But if you really wanna get one of the other two, having a look HERE, I'd take the 5700X non X3D cause better singlescore. But if this really translates to better performance in VaM, and having the 5700X3D as the only alternative, I dont know which one to pick really. I'd take a 5800X3D. Which is losing as well against a 5700X in singlescore 😂

5700X is a little beast actually.
 
HolySchmidt said:
Dude, just take a 5800X3D and you have the fastest you can get for AM4. No more worries.

But if you really wanna get one of the other two, having a look HERE, I'd take the 5700X non X3D cause better singlescore. But if this really translates to better performance in VaM, and having the 5700X3D as the only alternative, I dont know which one to pick really. I'd take a 5800X3D. Which is losing as well against a 5700X in singlescore 😂
Click to expand...
thanks for the advice, but if I had the finances to buy 5800x3d and cooling for it, I would switch to a new platform :)
 
Benchmark-20241116-034518.png
 
HolySchmidt said:
Faster than a 13900KS in average, but not really much. Min 1% have a huge improvement though. Thanks for testing!
Click to expand...
This one was using DDR4. The lower one was with faster RAM.

Benchmark Result Discussion

I want to try it, but I don't know how to configure it for my hardware. Can you help mi with that? I was searching for a tutorial/guide but I think there is none The guide is on the hub page that you get the software.
hub.virtamate.com hub.virtamate.com
I was like wth, these results were much worse than my 13900k [non KS] + 6000 Mhz RAM. I mean i'm at ~ 130fps 1%low at baseline 3 QQ
With good RAM they seems on tie pretty much overall, with worse physics time on intel
 
Useful data on the 9800x3d. Looking at my 5800x3d it's looking really hard to justify upgrading.
 
Yep, thats all i can squeeze outa this X3D on this gpu. Will have to wait for 5090 to open it up more.
Benchmark-20241116-215557.png
Benchmark-20241120-215555.png
 
Last edited:
Seraphim said:
Yep, thats all i can squeeze outa this X3D on this gpu. Will have to wait for 5090 to open it up more.View attachment 431447
Click to expand...
Thank you! :)
As I assumed amd did really well on maxing the out of box performance again.
Honestly I really love them for this, but on the other hand it makes most reviews literally useless..
Intel can only be evaluated correctly with custom profiles and faster ram, which they - almost - never use.

May I ask how was the temps and power? Is it managable on air easily, right?
 
Last edited:
Sween said:
Useful data on the 9800x3d. Looking at my 5800x3d it's looking really hard to justify upgrading.
Click to expand...
I am in the same boat with an ageing 6800 XT. I really want to buy something new n shiny, but I can still push most scenarios to 100ish fps with some extra tweaking.
Tough I may invest to blackwell .. if wont be too overpriced .. probably a bit later after release .. or just try grab a 4090 when folks switching to 5090!
 
Ultra 9 285k test.


Testing Platform:
4090, G.Skill 64Gb - 6400Mhz CL 32-39-39-102


Let's start with my 13900k. HT disabled, since it gave the best performance in VaM, and with latest microcodes.
Benchmark-13900k-no_patch.png
And with CPU patch [affinity=1,2,3,4,5,6,7,8]
Benchmark-13900k-patched.png

It's been a while since i used the benchmark... The results were worse than i had a few months back. Probably cuz of bios and microcodes updates QQ

285k
Same config for RAM, xmp at 6400Mhz.
Benchmark-285k-no_patch.png

And with patch [affinity=1,2,10,12,13,11,23,24]
Benchmark-285k-patched.png

/// EDIT 26.11.2024 //
It seems i messed up one P and E cores.
The correct affinity for 285k is:

affinity=1,2,11,12,13,14,23,24

Which leads me to a slightly low's improvments:
Benchmark-20241126-155037.png

I found some better RAM sticks btw, and waiting for a delivery in a day or two.
// EDIT 26.11.2024 ENDS //

So yeah... As expected, gaming performance is slightly worse on the new platform.
I might get closer to the old system, or even beat it, with faster RAM, but right now, CUDDIM sticks are terribly overpriced and... not really available on the UE market lol

So, does new intel sucks for VaM?
Yeah, probably.

Will it get better?
Maybe. It's brand new platform, and as you can see at my affinity settings for the CPU patch, P-cores are all around on the CPU.
Windows, and, or BIOS, updates might bring some more performance over time.

But, there are also some goodies.
Scripts times seems to be slightly reduced. Which proves the new CPU productivity improvements.
And temperatures.
It's almost 20 Celsius degrees colder now, using the very same cooler solution, at Simpler physics test [the hottest one :ROFLMAO: ].
13900k peak around 82 degrees, while hightest temp i got for 285k was 65 degrees.
 
Last edited:
trety said:
Ultra 9 285k test.


Testing Platform:
4090, G.Skill 64Gb - 6400Mhz CL 32-39-39-102


Let's start with my 13900k. HT disabled, since it gave the best performance in VaM, and with latest microcodes.
View attachment 434148
And with CPU patch [affinity=1,2,3,4,5,6,7,8]
View attachment 434159

It's been a while since i used the benchmark... The results were worse than i had a few months back. Probably cuz of bios and microcodes updates QQ

285k
Same config for RAM, xmp at 6400Mhz.
View attachment 434160

And with patch [affinity=1,2,10,12,13,11,23,24]
View attachment 434161

So yeah... As expected, gaming performance is slightly worse on the new platform.
I might get closer to the old system, or even beat it, with faster RAM, but right now, CUDDIM sticks are terribly overpriced and... not really available on the UE market lol

So, does new intel sucks for VaM?
Yeah, probably.

Will it get better?
Maybe. It's brand new platform, and as you can see at my affinity settings for the CPU patch, P-cores are all around on the CPU.
Windows, and, or BIOS, updates might bring some more performance over time.

But, there are also some goodies.
Scripts times seems to be slightly reduced. Which proves the new CPU productivity improvements.
And temperatures.
It's almost 20 Celsius degrees colder now, using the very same cooler solution, at Simpler physics test [the hottest one :ROFLMAO: ].
13900k peak around 82 degrees, while hightest temp i got for 285k was 65 degrees.
Click to expand...
Intel said that the new platform will be almost exactly the same performance as the last one, but at up to 50% less power usage. Ryzen 9950x3d might actually beat it once it's released in january
 
turtlebackgoofy said:
Intel said that the new platform will be almost exactly the same performance as the last one, but at up to 50% less power usage. Ryzen 9950x3d might actually beat it once it's released in january
Click to expand...
There is no point for AMD to beat it with 9950 x3d.
285k is right now even slightly slower than 13900k [2 years old Intel CPU], and totally crushed by 7800x3d in games.
There is no point to beat a dead body with a new x3d chips.

In short, 285k would be a great CPU, if it would be released 2 years ago. Now it's outdated as fck.

I'm gonna keep it, cuz i'm Intel fanboy, and performance is pretty much the same as for my last, 13900k, build. But it's a way more 'civilized' to work with in terms of power consumption and temperatures.
So it's like 'quality of life' upgrade. Nothing more.
 
trety said:
Ultra 9 285k test.


Testing Platform:
4090, G.Skill 64Gb - 6400Mhz CL 32-39-39-102


Let's start with my 13900k. HT disabled, since it gave the best performance in VaM, and with latest microcodes.
View attachment 434148
And with CPU patch [affinity=1,2,3,4,5,6,7,8]
View attachment 434159

It's been a while since i used the benchmark... The results were worse than i had a few months back. Probably cuz of bios and microcodes updates QQ

285k
Same config for RAM, xmp at 6400Mhz.
View attachment 434160

And with patch [affinity=1,2,10,12,13,11,23,24]
View attachment 434161

So yeah... As expected, gaming performance is slightly worse on the new platform.
I might get closer to the old system, or even beat it, with faster RAM, but right now, CUDDIM sticks are terribly overpriced and... not really available on the UE market lol

So, does new intel sucks for VaM?
Yeah, probably.

Will it get better?
Maybe. It's brand new platform, and as you can see at my affinity settings for the CPU patch, P-cores are all around on the CPU.
Windows, and, or BIOS, updates might bring some more performance over time.

But, there are also some goodies.
Scripts times seems to be slightly reduced. Which proves the new CPU productivity improvements.
And temperatures.
It's almost 20 Celsius degrees colder now, using the very same cooler solution, at Simpler physics test [the hottest one :ROFLMAO: ].
13900k peak around 82 degrees, while hightest temp i got for 285k was 65 degrees.
Click to expand...
Thank you for showing it!
Honestly I think it has indeed got more performance, look at the max fps numbers, they are consistently higher. Altough its obvious there are bottlenecks, hence the lows did not improve..
It is just a different beast again, I think intel somehow went from plug-n-play to need-to-tinker style over the years. Basically switched sides with amd..
If it gets the fastest possible ram and a proper overclock with probably a half year of post release updates its going to show its fangs and claws, I bet. But best case it can shine in subforums and topics such like this. Reviewers wont do the extra miles for it.
 
hmmm, what happened to my baseline 3 benchmark? very low, and I have fps 20-30 in many scenes

What do I need to update? CPU or more RAM?

When benchmarking the system says mostly 70-90% is GPU , RAM 50% > 90%
Screenshot_1.png
 
ldog said:
Why do most of you test in 1080p with mega gpu's? Why not test more in VR?
Click to expand...
Because we wanna test CPU performance. Therefore you have to minimize the GPU usage.

The VaM benchmark is a mix of CPU and GPU usage, and there's 1440p or 4K results more than enough for testing the GPU, we already know what the 4090 is capable of. But with new processors coming out we have to test them, and that's why.

VaM is special in general as there are many scenes/situations where you are CPU limited anyway, no matter what GPU.

@bbaoni
Most probably your CPU. If you are not running DDR4 3200 CL16 minimum, you could buy new RAM as well. 32GB is enough.
5800X3D should give your system a remarkable boost. That's the fastest you can buy for AM4 platform.
 
Last edited:
HolySchmidt said:
Because we wanna test CPU performance. Therefore you have to minimize the GPU usage.

The VaM benchmark is a mix of CPU and GPU usage, and there's 1440p or 4K results more than enough for testing the GPU, we already know what the 4090 is capable of. But with new processors coming out we have to test them, and that's why.

VaM is special in general as there are many scenes/situations where you are CPU limited anyway, no matter what GPU.

@bbaoni
Most probably your CPU. If you are not running DDR4 3200 CL16 minimum, you could buy new RAM as well. 32GB is enough.
5800X3D should give your system a remarkable boost. That's the fastest you can buy for AM4 platform.
Click to expand...
increasing resolution doesnt increase CPU load, but rendering same resolution in VR does. In VR unity has to render each frame twice for each eye, making twice the draw calls which cost CPU. So a CPU is kinda important for it to work well in VR.
 
turtlebackgoofy said:
increasing resolution doesnt increase CPU load
Click to expand...
I never said that.

As you may probably know, you are always CPU limited OR GPU limited. Never both at the same time. So, testing in VR, which increases the work on the GPU heavily (times 2), makes no sense if you wanna know what difference the CPU makes.
But, as I said, VaM is special. This is not your standard gaming-benchmark. You can easily provoke a CPU limit in every resolution, with every CPU, even in VR. Because of the physics calculation and the VaM engine.

Additionally, the VaM engine has a frame limit which was at 340 I think? So, testing in VR, with all the different resolutions the many headset provide, makes no sense cause it's not comparable anyway. Testing in 1080p, 1440p and/or 2160p is the best way cause everyone can relate to that. And even in 1080p you are (kind of) GPU and CPU limited. The 4090 was the first card reaching the engine limit and we were all like "what's going on here?".

If you wanna test in VR, do it. You can throw anything in here. Every benchmark is welcome. But for the best comparison use 1080p, that's the resolution we have the biggest set of data and you will be limited CPU and GPU wise anyway.
Of course we could all switch to 1440p for example but there's people in here still gaming in FHD. And, like I said, posting all different variations from their headsets plus maybe internal upscaling as well, this would be a total mess and we can't compare nothing really. And some people don't even game in VR.

Best way would be making a 1080p, 1440p and a 2160p run and if you own a headset bench that as well without up- or downscaling. So everyone can pick the data needed.
 
You must log in or register to reply here.

Similar threads

VaMRainey
December Holiday Event Discussion
Replies
2
Views
684
DateKnight
DateKnight
B
Setting the scene discussion
Replies
4
Views
854
booyakshaka
B
R
Benchmark - clear install vs many vars
Replies
19
Views
5K
ruinousdeity
ruinousdeity
gghhoosstt123
Unity engine fuckery
Replies
1
Views
833
dnc
D
G
VaM's cache: SSD vs HDD vs none - results of rudimentary testing
Replies
14
Views
3K
mikei9
M
Back
Top Bottom