CPUs, GPUs and benchmarking
Recently bought a GTX 1060 3GB GPU to use with an HTC Vive for VR games. The GPU worked well with an AMD FX-8130 CPU and 8 GB ram. Most games needed the settings to be turned down nearly to their lowest settings to allow the game to run at 90+fps so there would be no hickups in the video in the VR headset.
Because of this limitation, I was keeping an eye out for other GPUs, and came across a good deal on a GTX980Ti. Even thought this is one generation old (Maxwell architecture, vs the new Pascal which is what the GTX1060 card is).
I bought the card, brought it home, installed it, and saw virtually zero performance improvement in the games I tried. Imagine the disappointment. I just spent a few hundred $$ to gain zero performance 🙁
Time to investigate and runs some bench marking tools. I’ve never been one to run benchmarking tools, as I never understood the use for them other then just to document your system’s performance, but I’ve learned a lot in the last couple of days by running the PassMark benchmarking tool, and looking at benchmarks that other people have done on their systems.
My system didn’t improve because in it’s current state, the GPU performance is bound by my CPU’s performance. Both the GTX1060 and GTX980Ti scored the same 3D performance because the CPU couldn’t keep the GPU pipeline full of data to process.
Looking at the PassMark database (the database I could only find by using the PassMark performance test software, and not on their website) for users of GTX1060 GPUs I can see that others with similar CPUs were getting similar 3D benchmarking scores to me, while others with faster CPUs were getting faster (much faster) 3D benchmarking scores. The story extended to the GTX980Ti GPU and CPU data in the database.
In the current state of gaming (DirectX 11 and below) single core performance is what is desired. DirectX 11 (and older) don’t take advantage of multiple cores, so even though I have an 8 core CPU, one core is not that fast. There are CPUs that are double the speed per core, and even though they may only have 4 cores, they will run the same game at a much higher frame rate.
DirectX 12 will utilize multiple cores so take advantage of multicore systems, but for now, most games are still DX11.
Looking at more data, upgrading the CPU to a slightly better one to use with the GTX1060 seems like will provide more 3d performance then doing the same upgrade on the GTX980Ti.
So what all this boils down to is that I need a new CPU+motherboard to take advantage of either the GTX1060 or the GTX980Ti.
As luck would have it, I’ve been putting off upgrading my main workstation for quite a while now and this is a good excuse to make the jump to a latest generation i7 intel CPU with top notch performance for both CAD, design, development and some gaming and paired with the GTX980Ti, I should see a huge improvement in 3d performance.