Pascal architecture is much slower than Maxwell...

Hi Justin,

In fact Pascal architecture is much slower than Maxwell. Could you please look at my benchmarks down below? The point is to measure old architecture GPUs with optimized DNN libraries for that specific architecture - CUDA 7.5 cuDNN v4.0.

Measured in CNTK, Caffe under Windows 7 and Ubuntu OS.

GTX 980 Ti - CUDA 7.5 with cuDNN 4.0
Caffe Performance: 5242 imgs/s

GTX 980 Ti - CUDA 8.0 RC with cuDNN 5.1
Caffe Performance: 4183 imgs/s

GTX 1080 - CUDA 7.5 with cuDNN 4.0
Caffe Performance: Not Applicable

GTX 1080 - CUDA 8.0 RC with cuDNN 5.1
Caffe Performance: 4628 imgs/s

Best Regards,
Ondrej

Start by reviewing the reported CNTK and Caffe benchmark results for GTX 980 Ti and GTX 1080 across CUDA 7.5/8.0 and cuDNN 4.0/5.1 on Windows 7 and Ubuntu. Done means reproducing or explaining the Pascal-versus-Maxwell performance discrepancy with the relevant benchmark configuration documented.

Written by the indexing model from the issue text.