Futuremark Announces VRMark, the Virtual Reality Benchmark

According to reddit the orange test is pretty CPU limited with anything above a 1070. My 5820k + Titan XP is getting basically the same scores as a 5830k + 1070
 
TheSmJ said:
For what reason?

I've never bought a copy of 3DMark until the price hits $5, but if nothing else it's nice to see how your system is performing to other systems of the same configuration using a 100% repeatable test loop.
Click to expand...
Because it is not worth the money?

But it only tells you how it compares again other systems IN 3DMARK! That is it.

If that is important to you, by all means, get you some.
 
TheSmJ said:
Not sure I understand the source of your hostility, but thanks for answering.
Click to expand...
Not hostile about anything, the point is that benchmark only tells you how well it runs that benchmark. It is not a valid test to tell you anything about how particular VR games will run.
 
Just ran the demo on my system, system specs in sig. What do you guys think?

dI15QSG.jpg
 
bizzmeister said:
Just ran the demo on my system, system specs in sig. What do you guys think?
Click to expand...

First off, I think that is laughable and simply bad to suggest that.

Have you looked at any of these articles?

If you want to play "kiddie" games in VR. You are set. I guess VRMark is about cartoon games.... If you want to play games that I think the majority of [H] readers want to play, the GPU will not get you there. CPU is fine for now at that high of a clock. GTX 1060 is the bottom rung that I would personally purchase for VR now. GTX 1070 is my suggestion. With the 1060, you are going to have to give us some fidelity, the 1070 will do a LOT better in terms of IQ.
 
Shintai said:
https://www.overclock3d.net/reviews/gpu_displays/vrmark_gpu_performance_review

03142501626l.jpg

03142257541l.jpg
Click to expand...

How does a website that has NEVER done any proper VR performance testing tell me a tool is useful? LOL!

"It is immediately apparent that VRMARK is a useful tool for judging the VR readiness of a PC, with the ability to simulate a VR workload without a headset and then provide results that will allow PC enthusiasts to differentiate the VR performance of many different systems and various competing GPUs. "

Hey look, 6,971 points! It is apparent this tool is incredibly useful. What a POS. I gotta give 3DMark credit though, they are one of the few companies in the world that can bilk gamers and get away with it over and over and over.
 
Activate: AMD said:
To play devils advocate here, do the results not at least track with yours? I agree that the be-all/end-all really should be "real world" testing, but abstraction of the results for direct comparison isn't a totally invalid way of gaining some insight. You have done something very similar yourself with your little VR leaderboard, the only difference is that 3DMark has taken it a step further and created some kind of weighted score rather than some average of very disparate results. If used correctly 3DMark is just a tool to validate other data, just one piece of a big puzzle. The danger is people not thinking beyond the numbers, but thats an issue with everything.
Click to expand...
Actually they track very close to what we have on the VR Leaderboard, with the exception of the 1070 and 980Ti being flip flopped. Scaling can change greatly from game to game as well. However, look at what it is telling that VR Ready and Oculus Ready is. I would suggest that any gamer that sees his GTX 780 system is "VR Ready" and dumps the money into it, will be tremendously disappointed.

1478229895QfdsCZwbbR_6_2.png
 
And it takes into no account that the programs to run VR actually work....
 
pj- said:
I don't necessarily think this test is useful, but it does highlight something that I think your tests lack. CPU testing.
Click to expand...
And I would suggest that most HardOCP readers are in no way CPU limited for the most part which is why I have not spent a lot of time focussing on it. The fact of the matter is that you are tremendously more likely to be GPU limited in VR gaming, the same as desktop gaming.

A 4.0GHz CPU 4C CPU is not going to be holding you back in any way in terms of Reprojection from what I have seen.

Considering your statement, let me know what CPU limited VR title you would like me to look at I will give it consideration.
 
Then start downclocking your CPU and limiting threads and let me know what your hard work and time show. I have other things to do.
 
In general, 3DMark I've found useful for comparing own system variations when OCing etc. Then again any can benchmark can also be used but 3dmark does have things more broken up. For example OCing memory affects Physics processing and you can find the best memory speed and memory settings to maximize that. Sometimes the fastest speed is not always the best setting, a slower speed with faster timings can win out. CPU likewise, max clock may not always be the fastest if you have to lower other settings like memory, bus speed etc. In short it can be a useful tool. Now does it tell you how a certain or any game will play? Not at all from my experience.

Another use I like 3DMark for is to use Prime95 and 3DMark for stability test - if you can run both for hours your system is super stable. Nothing comes close I've found.

Now if the VR test validates or is very consistent with real world data from HardOCP then that does indicate it has a level of accuracy there. Than again maybe FutureMark just made sure VRMark reflected HardOCP hard earned data closely :D
 
You must log in or register to reply here.
Back
Top