Multi-Res Shading, Closing Thoughts

Arizona Sunshine is available on both Oculus and Steam for the Rift and Vive, but our testing was done through SteamVR. It takes place in a post-apocalyptic world where you can handle motion-controlled weapons to you know…shoot zombies.

Click to Enlarge

Click to Enlarge

We see an interesting result here. The GTX 1060 results in a higher frame rate, both in delivered and unconstrained frame rate, but if you look at the green/red line graph you'll see that the NVIDIA card is, in fact, running at a less consistent frame time than the RX 480. In a normal gaming scenario that would be more concerning, but here, as long as the frametimes are completed in enough time to meet the necessary refresh rate, then the VR experience shouldn't be affected. Looking at the interval graphs for both cards I see an interesting comparison: while the RX 480 has sustained blocks of time with 45 FPS effective gameplay, the GTX 1060 has more individual spikes down, indicating less grouped dropped frames. Overall, the GTX 1060 has a 9% edge.

Below you'll find results from Raw Data, one of the first successful games on the HTC Vive. While the game itself is solid we are not comparing the GTX 1060 to the RX 480 in this case. Instead, I want to use it to demonstrate the capability of MRS, multi-res shading. Launched with the GeForce GTX 980 Ti in 2015, MRS adjusts the rendered resolution of the image to better match the effective resolution of the image as it is scaled through the lens in your headset. 

NVIDIA's solution is Multi-res Shading that can divide the image that the game engine wants to display into nine different viewports.  The center viewport, the one the user will 99% of the time be focused on and the one without lost pixels to the warping, remains the same resolution and maintains the detail required for a great gaming experience. However, the surrounding viewports can be adjusted and sized to more closely match the final warped resolution that they will display at in the VR headset. The image still has to go through a final warp to the correct shape for the VR lens but the amount of data "lost" along the edges is minimized and thus performance can be improved for the gamer.

Raw Data supports MRS at settings of 0 (off), 1 (normal) and 2 (high). 

Click to Enlarge

Click to Enlarge

Let's look at what is happening with the GeForce GTX 1060 running at Medium settings. The average frametime and frame rate improve, though only with the MRS setting of 2. MRS 0 and MRS 1 seem to offer very little improvement with this GPU, but MRS 2 results in a 13% performance improvement for little to no degradation in image quality.

What about the GTX 1080 that has a bit more headroom?

Click to Enlarge

Click to Enlarge

While all three settings allow the GTX 1080 to run at Medium quality settings with very few dropped frames, the unconstrained frame rate increases by 5% at MRS 1 and by 18% at MRS 2. This added performance capability could very well be the difference between being able to run a game at Medium settings or High and we have witnessed other games with MRS support show similar advantages. More testing to be done next week!

Initial Closing Thoughts

At this point, I feel confident that FCAT VR is a fantastic tool for reviewers, developers, and end users to take advantage of and measure all aspects of VR gaming. Much like the use of FRAPS democratized basic performance analysis for forums and communities, keeping the public engaged and knowledgeable, I think FCAT VR can do the same thing for virtual reality. It’s not quite as dead simple, and there are still features I would like to see added to the software (easy including of 95th / 99th percentile frametimes), but it's easily the best and most robust tool available.

Add to it that NVIDIA is going to open source the code for analysis, and I fully expect other programmers to take a stab and developing tools to deal with the data files and include even more features.

Our first results, though not covering nearly as many games as I would like at this point, show the GeForce GTX 1060 6GB card coming out ahead of the Radeon RX 480 8GB. Chronos, Dirt Rally, and Obduction on the Oculus Rift all indicate significant performance advantages for the NVIDIA card in this price range, despite the claims of VR dominance by the Radeon marketing and branding teams. Edge of Nowhere shows a very small edge for the GTX 1060, effectively creating a tie.

The battle between the GeForce GTX 1080 and the Radeon Fury X only has one data point currently, but honestly based on the last several weeks of testing, the battle isn’t worth fighting. AMD does not have a competitive part in the high-end consumer graphics space for standard PC gaming and that doesn’t get any better in VR. In fact, it gets worse. I said it in my GeForce GTX 1080 Ti review, but AMD needs Vega and it needs it soon.

From me, expect to see much more VR performance evaluation soon, including in upcoming graphics card reviews as the spring and summer hit. If you have a VR headset of your own, head over to and download the tool yourself – I’d love to see your thoughts!

« PreviousNext »