Rift Compatibility Tool

When I look at the specs of the laptop you are linking to I see both a mobile CPU and a mobile GPU. Those are in general a lot slower than their desktop brethren.

When I do a quick google on the performance differences I find that the recommended nVidia GPU (GTX970) is 94% faster than your GTX 880M...

http://gpu.userbenchmark.com/Compare/Nv ... 577vsm8540

Thanks!

Besides VR, I now feel that this laptop was a total waste of 3K....

Don't bother with that compatibility tool. Run the 3DMark firestrike demo from steam. If you score over 9200 your good to go

"laminblake" wrote:
Don't bother with that compatibility tool. Run the 3DMark firestrike demo from steam. If you score over 9200 your good to go

Many high end laptops with Optimus are not supported for technical reasons, whatever 3DMark score they may have.

Palmer confirmed on Reddit that they will not block people with weaker hardware from using the Rift.

"Myrskalle" wrote:
Palmer confirmed on Reddit that they will not block people with weaker hardware from using the Rift.

That's definitely a plus... That gives people a chance to experience it, judge their system's performance and the decide if and when they choose to upgrade...

Kudos!

Also, do you suspect that programs such as watching a movie, etc will be less GPU intensive than playing the newest game on the rift, for example....? I would assume so....

The Compatibility Tool is not perfect, but it is fairly accurate in most cases. The test is based on a very large range of different common hardware (both whitelisted and blacklisted) though it's possible you could have something uncommon that's not detected. I would take the results seriously unless you really know you have good hardware and there is a mistake.

Even if you have slower hardware, you won't necessarily be blocked but the experience can be poor and uncomfortable.

@Mattborkin: Laptops mobile GPUs are not supported at all, so that's probably why you've failed. Even the best mobile chips are significantly slower than desktop chips.

@laminblake: What you are saying is incorrect, and is part of the problem of why people are confused. Getting 9000 in 3DMark DOES NOT mean your PC is ready for Rift. First, mobile chips don't work (for example, with Optimus), so getting 9000 on a laptop means nothing since laptops don't work. Then there is SLI, which is also not supported. If you get a high score with 2 (or 3) cheaper cards, it won't work cause SLI doesn't work. The lump score also can hide deficiencies in your machine, for example if you have a monster GPU but cheap CPU (or vice versa) where you get a decent total score but certain components aren't good enough.

Thanks, @cybereality....

I now have a $2,500 budget to build a new rig - so that's my new project.

"cybereality" wrote:
The lump score also can hide deficiencies in your machine, for example if you have a monster GPU but cheap CPU (or vice versa) where you get a decent total score but certain components aren't good enough.

Do you think that VR-focused benchmarking tools (like Futuremark's upcoming VRMark) would give a more reasonable comparison, or is it really down to individual applications and to "stick to the spec" when discussing potentially lesser hardware? Not asking about VRMark's usefulness specifically, but the general sentiment about benchmarks for VR, since unlike most 2D applications performance is much more tightly linked to usability and comfort.

Maybe VR benchmarking should only be used to compare min spec and higher builds? So it'd be more about future-proofing and reaching max settings, and not for answering "can I run this." (I completely get why the only *real* answer to "can I run this" is to go with the official recommendations, hence the official Compatibility Tool -- the platform holder can't and shouldn't make any promises about hardware below min spec, it's up to developers if they want to advertise support for lower-end machines. Which...would probably not be helpful in the long run anyway.)


Unrelated: kudos to the team for identifying the USB controller/driver compatibility issues and building that into the tool, I bet that caused a lot of headaches internally. High fives to the QA and devs who got in front of that particular bus. And my sympathies to the Support people who'll take the hit when some people inevitably didn't run the tool and can't get it to work. It's not a topic I saw discussed publicly before the tool was available, any idea if that problem was seen with the DK2? Or did the custom hardware (as opposed to some presumably off-the-shelf components in the DK2) and higher throughput of Crescent Bay/CV1 introduce new, more stringent USB 3.0 requirements?

Sure, a dedicated VR benchmark app could be useful. However, you do run into issues that vary depending on the engine used, content complexity, where the bottlenecks are, etc. You can see this in game benchmarks, where an Nvidia card will out-perform a comparable AMD card, and then another game will score better on the same AMD card. So which is the better card? You never know.

At least with the Compatibility Tool, it's looking for specific hardware that Oculus has decided will be adequate. It is an arbitrary line drawn in sand, but it's the best option at this point. Developers will all be using those recommended specifications to build their games, and the games will have to be optimized to run well on that level of hardware.

In any case, surely people in the community will do their own benchmarks with real hardware and real games once Rift releases and people can come to their own conclusions.

The USB 3.0 issues are unique to the consumer Rift, since it's really pushing the limits of what's possible. What the issue boils down to is that the USB 3.0 specification was ambiguous in places, and certain vendors may have interpreted the spec in different, conflicting, ways. Not wrong, just different, but these differences lead to undefined behavior. You don't notice it with simple devices (like a thumb drive or whatever) but with all the complexity of Rift, it becomes an issue.