Benchmark Result Discussion

[TToby]

@endolu
Thank you very much. The XMP profile was the first thing I have checked, because this is the only thing I really know.
I always was too lazy to look after all of those hundreds of BIOS settings. Did my last OC approx 10 years ago.
I just have clicked on those "Automatic OC" buttons in the ASUS Bios for some kind of optimization.
In addition to this, I have installed most of the bloatware coming with the board like Armory Crate for the bling-bling and AI Suite3, wich has some sort of build in "CPU optimization" tools, aso. I think I should do a spring vleaning soon.
My Windows Installation is very old, too. Maybe 7 years without re-installation.
Thank you for pointing me to the CPU! I guess I have seen in CPU-z that my 5800 is running much lower than yours.
This is definitely the next thing I have to look after.
Though, my stock AMD 6900xt is running very OK at approx 2500MHz. at 70-80 C max with watercooling and two radiators.
I am very happy to have the benchmark for testing this out. VaM is almost the only "game" I am running for a long time, so this is somewhat important to me.

 

[Jiraiya]

Here are my desktop results, still substantially lower than the majority being posted so that counts the v-sync with Oculus out as a problem.

Again, I am confused because if you look at the "total" times in ms for each test, mine are lower than those shown in rernat's tests.
For clothsim I get a total time of 5.54ms while rernat got 10.83.
Yet, his average fps is 85.65 and mine is 83.36.

How can my average time per frame be lower, but the FPS also be lower? Surely if it takes less time to do each frame it should be a higher fps! Something very strange is happening.

 

[halopcdraco]

Well it seems that I have the worst setup here. My PC isn't overclocked in any way. For some reason, the more .VAR files I download, the slower VaM loads. I used to be able to have 2 or 3 people. Now it lags hard with 2 or more people. One quick question. Am I suppose to unpack all .VAR files or load them as they are?

 

[rernat]

Am I suppose to unpack all .VAR files or load them as they are?

Load them as they are. Using a fast SSD-Drive will help alot.

 

[MacGruber]

Again, I am confused because if you look at the "total" times in ms for each test, mine are lower than those shown in rernat's tests.
For clothsim I get a total time of 5.54ms while rernat got 10.83.
Yet, his average fps is 85.65 and mine is 83.36.

Not 100% sure....performance is complicated as you probably noticed by now....but I think its because @rernat got a first generation Ryzen 1500X, yours is a 5th generation 5600X that has a 50% faster singlecore speed. RAM is probably also DDR3 vs. DDR4? So, my assumption would be that rernat is limited by CPU, while you are limited by GPU.

 

[halopcdraco]

Load them as they are. Using a fast SSD-Drive will help alot.

I have VaM on my fastest ssd out of the 4 I have. Samsung 870 evo.

 

[Jiraiya]

Not 100% sure....performance is complicated as you probably noticed by now....but I think its because @rernat got a first generation Ryzen 1500X, yours is a 5th generation 5600X that has a 50% faster singlecore speed. RAM is probably also DDR3 vs. DDR4? So, my assumption would be that rernat is limited by CPU, while you are limited by GPU.

I think you missed my point of confusion. I am willing to accept that there could be hardware reasons his might be faster (for whatever reason) however it's slower according to the benchmark. The time taken in ms for all listed tasks is longer with a total ms per frame being longer than my system. If it takes longer, how on earth are his FPS higher than mine?

 

[MacGruber]

If it takes longer, how on earth are his FPS higher than mine?

Not sure either. The time measurements show are those measured by VaM the same the VaM's build-in "Performance Monitor" feature displays. Maybe that does only account for a part of the render time?

 

[danood]

xmp to 3200mhz ram.

 

[rernat]

I have VaM on my fastest ssd out of the 4 I have. Samsung 870 evo.

This is only a SATA III SSD (max around 600 MB/s). Consider using a M.2 on PCIe which allows almost 10 times higher transfer rates.
e.g. : Seagate FireCuda 530 2TB on PCIe 4.0 M.2 slot with 6.342 MByte/s read and 6.414 MByte/s write.

 

[DarkFantasy]

This is really sweet! I am heading out for a week but have been playing with it as I have time in Desktop mode with different things enabled and disabled. Its interesting how much the AV+Mega+VPN impact FPS. Here are two tests in Desktop mode, (running on a mounted and encrypted drive on an 3d NAND NVME PCIE M.2 SSD). When I get back and have time I will play with VR and other background processes. After seeing others I think I should be getting more.

Desktop mode with Trend+Mega+VPN:

Desktop mode with Trend+Mega+VPN disabled. The average is close to the same but the difference in the max peak is huge...

 

[Bob Nothing]

Thanks @MacGruber for this, this is something I think the community has wanted/needed for a long time. So, I just got this iBuyPower PC (normally I build but I got sick of not being able to buy a GPU, so I bought this). It's an i9 with a 3090. I really feel like it should have performed better, so if anyone notices anything glaringly wrong in my config, please let me know.

 

[Bob Nothing]

xmp to 3200mhz ram.

Holy crap??? How in the world did you get results that high?

 

[RandomVAMUser]

as desktop only user...here is my potato result

important notes? ...umm (since basic info doesn't show it) i have/use:
-2x a2000 nvme (gen3)
-2x8 hyperx fury xmp 3200/CL16 (there's also 3000/CL15 profile, didn't bother testing latency difference, also ryzen prefers higher clock)
-pbo force on (quet-mode off), infinity fabric at 1600
-resizable bar/sam on (tbh didn't noticed any difference..than again i always keep gpu locked at 60frames with RTSS, vsync/freesync force off)

Spoiler: sidenote: too easy to cheat

after some testing: even with all verifications there's still ways to "officially" rig results

gotta keep that e-peen up? ?

 

[MacGruber]

sidenote: too easy to cheat

Verification only protects against accidental issues caused by noob users. Can't prevent intentional manipulation.

 

[rernat]

sidenote: too easy to cheat
after some testing: even with all verifications there's still ways to "officially" rig results

?

Why the hell should you cheat that benchmark. If you need a bigger cock get into vam and create one. ?

 

[geo_gan]

Thanks @MacGruber for this, this is something I think the community has wanted/needed for a long time. So, I just got this iBuyPower PC (normally I build but I got sick of not being able to buy a GPU, so I bought this). It's an i9 with a 3090. I really feel like it should have performed better, so if anyone notices anything glaringly wrong in my config, please let me know.

Turn off 2x scale! Just run at 1x - that 2x VR resolution on a CV1 is crazy

Question for all - how did you enter name in results picture? Mine just says Anonymous!

 

[Smith]

Question for all - how did you enter name in results picture? Mine just says Anonymous!

It's Your Creator Name

 

[vamurai]

Wow... so many very nice systems out there. If y'all wanted to see what this badboy runs like on a slow rig in desktop mode, step right up!

 

[Voithe]

Note: Updated my post with VR results.

 

[TToby]

@MacGruber You should add a warning message "This tool has the potential to make you mad"!

I am still optimizing my PC for more than 2 days now, because of Endolus much higher results with the same hardware. {mad laughing}. I have run the tool for approximately 30 times now... lalalalala ?
Today I will do a clean Windows installation.

 

[MacGruber]

I am still optimizing my PC for more than 2 days now, because of Endolus much higher results with the same hardware. {mad laughing}. I have run the tool for approximately 30 times now... lalalalala ?
Today I will do a clean Windows installation.

Note that I just added a section "My system seems to be performing slower than expected (...) What should I check? " to the Performance FAQ
Check it here: https://hub.virtamate.com/resources/benchmark.11336/

 

[Hedgepig]

Thanks for the test!

I bought a 3060 because I was warned that VAM's CPU bottleneck would slow down much faster graphics cards, hence they were not really worth buying until VAM 2 is released. I hope I made the right decision.

.

 

[Jiraiya]

@Hedgepig, I have a 5600X and a 3060 (no Ti for me .. boo) and was getting lower FPS than you in desktop mode, so it seems the Ti part certainly makes a difference, I think the 3070 and 3080 people also get better performance. I don't think the CPU is a "bottleneck" as such. Think of it more as a half and half. Speeding either half up certainly speeds the whole result, but you get diminishing returns if you don't reduce each half sort of equally. An awful CPU offsets a LOT of expense on a high end GPU where you would get better results spending less money on a slightly better CPU (if I explained that well, sorry otherwise).
So think of it less as a bottleneck and more of a balancing act with cost as a major factor.

 

[HolySchmidt]

@TToby
Don't come back until you got nearly the same results ? Freeze it!

I once owned a 1700x for a short time and was playing around in the BIOS. There's this one setting which endolu already mentioned I think that is responsible for core-threading or something like that. Where you first think of course this has to be on! After turning it off I had better game-FPS and this was also explained in some hardware-newspapers.

What I wanna say is: AMD is pretty sensitive when it comes to settings. Intel is more plug and play. You can get alot of FPS by just tweaking this and that.

I also owned the 6800XT for some days as you may already know. I ended up turning most of the AMD features OFF in the driver-software as the best option.

I suggest to install MSI Afterburner + Riva Tuner statistics server (included) with which you can easily monitor your FPS in every game.