SANDRA's GPGPU Image Processing benchmark runs through an array of filters on its reference data and offers up an aggregate score, derived from a number of individual results. Its GPGPU Cryptography benchmark churns through an assortment of workloads, and presents individual results for overall bandwidth, AES256
encryption and decryption, and SHA2-256 hashing bandwidth. CUDA and OpenCL code paths are available in these tests; we used both on the NVIDIA-powered cards and OpenCL on the
Radeon Pro WX cards.
| | SiSoft SANDRA 2016 SP2 | | Cryptography And Image Processing OpenCL / CUDA Performance Tests |
|
The Quadro P6000 takes the lead, regardless of whether or not CUDA or
OpenCL was used. The Radeon Pro WX 7100 puts up one heck of a fight though, and actually outpaces the pricier Quadro P5000. Also of note is the improved performance when using CUDA. There are fewer abstraction layers with
CUDA -- i.e. it runs "closer to the metal" -- hence the somewhat better performance.
The new Quadros put up some big numbers in the various GPGPU cryptography tests. The Quadro P6000 in particular jumps out to a big lead, thanks to its relatively large memory complements of increased number of cores.
