Dual NVIDIA Titan RTX Review Compute with NVLink Performance

10

Dual NVIDIA Titan RTX NVLink Rendering Related Benchmarks

Next, we wanted to get a sense of the rendering performance of using two NVIDIA Titan RTX GPUs with NVLink.

Arion v2.5

Arion Benchmark is a standalone render benchmark based on the commercially available Arion render software from RandomControl. The benchmark is GPU-accelerated using NVIDIA CUDA. However, it’s unique in that it can run on both NVIDIA GPUs and CPUs.

Download the Arion Benchmark from here. First-time users will have to register to download the benchmark.

2x NVIDIA Titan RTX NVLink Arion
2x NVIDIA Titan RTX NVLink Arion

As we will see in this section using dual NVIDIA Titan RTX GPUs with NVLink has a massive impact on rendering applications that we tested. Arion results effectively double, showing excellent scaling.

MAXON Cinema4D 3D

ProRender is an OpenCL based GPU renderer which is available in MAXON’s Cinema4D 3D animation software. A fully functional 42-day trial version is available for downloaded from the MAXON website here. Note: Even after expiration, the trial can still be used to measure render times.

2x NVIDIA Titan RTX NVLink Cinema4D
2x NVIDIA Titan RTX NVLink Cinema4D

In Cinema4D R20, the dual NVIDIA Titan RTX GPUs with NVLink cut our render times in half over the single GPU configuration. This shows excellent scaling.

OctaneRender 4

OctaneRender from Otoy is an unbiased GPU renderer using the CUDA API. The latest release, OctaneRender 4, introduces support for out of core geometry. Octane is available as a standalone rendering application, and a demo version is available for downloaded from the OTOY website here.

2x NVIDIA Titan RTX NVLink OctaneRender
2x NVIDIA Titan RTX NVLink OctaneRender

If you want to go fast here, the dual NVIDIA Titan RTX GPUs with NVLink setup simply blows away the other configurations we have tested with OctaneRender.

Redshift v2.6.32

Redshift is GPU-accelerated renderer with production-quality. A demo version of this benchmark can be found here.

2x NVIDIA Titan RTX NVLink Redshift
2x NVIDIA Titan RTX NVLink Redshift

Performance of the dual NVIDIA Titan RTX with NVLink setup is at the top of our results. Redshift is a very demanding benchmark, and these GPUs scale very well. We did not quite see half the rendering time of the single Titan RTX solution, but the dual configuration is not far off.

On our rendering benchmarks, one thing is clear: if you work against deadlines and need faster render times to get your job done and iterate faster, the dual NVIDIA Titan RTX setup performs extremely well in this domain.

Next, we will have 3DMark results before moving onto power consumption, thermals, and our final thoughts.

10 COMMENTS

  1. Incredible ! The tandem operates at 10x the performance of the best K5200 ! This is a must have for every computer laboratory that wishes to be up to date allowing team members or students to render in minutes what would take hours or days ! I hear Dr Cray sayin ” Yes more speed! “

  2. This test would make more sense if the benchmarks were also run with 2 Titan RTX but WITHOUT NVlink connected. Then you’d understand better whether your app is actually getting any benefit from it. NVLink can degrade performance in applications that are not tuned to take advantage of it. (meaning 2 GPUs will be better than 2+NVLink in some situations)

  3. Great review yes – thanks !
    2x 2080 Ti would be nice for a comparison. Benchmarks not constrained by memory size would show similar performance to 2x Titan at half the cost.
    It would also be interesting to see CPU usage for some of the benchmarks. I have seen GPUs being held back by single threaded Python performance for some ML workloads on occasion. Have you checked for CPU bottlenecks during testing? This is a potential explanation for some benchmarks not scaling as expected.

  4. Literally any amd GPU loose even compared to the slowest RTX card in 90% of test…In int32 int64 they don’t even deserve to be on chart

  5. @Lucusta
    Yep the Radeon VII really shines in this test. The $700 Radeon VII iis only 10% faster than the $4,000 Quadro RTX 6000 in programs like davinci resolve. It’s a horrible card.

  6. @Misha
    A Useless comparison, a pro card vs a not pro in a generic gpgpu program (no viewport so why don’t you say rtx 2080?)… The new Vega VII is compable to rtx quadro 4000 1000$ single slot! (pudget review)…In compute Vega 2 win, in viewport / specviewperf it looses…

  7. @Lucusta
    MI50 ~ Radeon VII and there is also a MI60.
    Radeon VII(15 fps) beats the Quadro RTX 8000(10 fps) with 8k in Resolve by 50% when doing NR(quadro RTX4000 does 8 fps).
    Most if not all benchmarking programs for CPU and GPU are more or less useless, test real programs.
    That’s how Puget does it and Tomshardware is also pretty good in testing with real programs.
    Benchmark programs are for gamers or just being the highest on the internet in some kind of benchmark.

  8. You critique that many benchmarks did not show the power of nvlink and using pooled memory by using the two cards in tandem. But why did you not choose those benchmarks and even more important, why did you not set up your tensorflow and pytorch test bench to actually showcase the difference between nvlink and one without?

    It’s a disappointing review in my opinion because you set our a premise and did not even test the premise hence the test was quite useless.

    Here my suggeation: set up a deep learning training and inference test bench that displays actual gpu memory usage, the difference in performance when using nvlink bridges and without, performance when two cards are used in parallel (equally distributed workloads) vs allocating a specific workload within the same model to one gpu and another workload in the same model to the other gpu by utilizing pooled memory.

    This is a very lazy review in that you just ran a few canned benchmark suites over different gpu, hence the rest results are equally boring. It’s a fine review for rendering folks but it’s a very disappointing review for deep learning people.

    I think you can do better than that. Pytorch and tensorflow have some very simple ways to allocate workloads to specific gpus. It’s not that hard and does not require a PhD.

LEAVE A REPLY

Please enter your comment!
Please enter your name here