These benchmarks were run using Iray 2017.1.2. Cards represented are of the Kepler, Maxwell, Pascal and Volta architectures. For cards with the same architecture which we have not listed you can usually extrapolate a reasonable estimate from the specifications.
These results are for Iray Photoreal. Note that for RTX, Iray 2017.1.2 does not utilise the new RT Core raytracing technology. We are currently rerunning our benchmarks in Iray RTX 2019.1.1. As such the performance data for RTX cards shown here is purely for CUDA core usage. See our article RTX Performance Explained for additional details on how RTX affects performance. For more details see the testing methodology below the results.
An increasing number of cloud providers are offering on-demand GPU resources. This is great news for users of Iray enabled applications and to give you an idea of what performance to expect relative to bare metal hardware you might find in your own computers we have run all of the tests for you. These tests are all GPU only, in some cases you may obtain additional performance by enabling the CPU as well but in our experience those resources are better left for other tasks.
For Google Compute Engine, rather than a pre-configured machine type consisting of a CPU and GPU, you attach GPU types to any of the supported machine types. This means you can actually run larger numbers of GPUs on machines with much less CPU resource than normal. Note however that you must have at least as many CPU cores as you have physical GPU chips to get full use out of all of the GPUs.
The DGX-1 and Quadro VCA are an appliance offering which has been specifically tuned to run demanding GPU accelerated applications. In addition to very high performance, the Quadro VCA specifically offers Iray IQ mode allowing large numbers of Quadro VCA appliances to be interconnected for extreme performance. We didn’t have access to large quantities of Quadro VCA appliances for our tests, however below are the results for a single Quadro VCA to give you a feeling for the performance as well as a single DGX-1. We put this in its own section because including it in the main bare metal graph scales the single cards too small!
All benchmarks have been performed under Linux (usually CentOS 7.4) with the latest available NVIDIA drivers (as of writing 384.98). Some tests were performed on older drivers where administrative access to the machines was not available. Where we have full control over the environment the following setup was utilised.
|Operating System||CentOS Linux 7.4.1708|
|Linux Kernel Version||3.10.0-514.10.2|
|NVIDIA Driver Version||384.98|
|Iray Version||Iray 2017.1.2 build 296300.3713|
|CPU||Intel Core i5 6500|
|PCIe||v3.0 Full 16 Lanes|
In order to ensure we are testing raw Iray performance we have developed a stand-alone benchmark tool based on Iray. Our tool renders a fixed scene multiple times and averages the results to ensure consistency. To ensure the results mean something for real-world use we utilise a non-trivial test scene, ensuring the GPUs have plenty of work to do. The image above is a fully converged version of our test scene.