On 05/07/2017 10:05, Andreas Beckmann wrote:
I had seen them, too, but haven't investigated them at all :-)
Should I upload the version we have right now? Is there an upstream
releasse announcement? What changed?
I found this [1]:
cuBLAS 8.0.61.2 New Features
cuBLAS Library
This update contains performance enhancements and bug-fixes to the
cuBLAS library in CUDA Toolkit 8. Deep Learning applications based on
Recurrent Neural Networks (RNNs) and Fully Connected Networks (FCNs)
will benefit from new GEMM kernels and improved heuristics in this release.
This update supports the x86_64 architecture on Linux, Windows, and Mac
OS operating systems, and the ppc64le architecture on Linux only.
The highlights of this update are as follows:
‣ Performance enhancements for GEMM matrices used in speech and natural
language processing
‣ Integration of OpenAI GEMM kernels
‣ Improved GEMM heuristics to select optimized algorithms for given
input sizes
‣ Heuristic fixes for batched GEMMs
‣ GEMM performance bug fixes for Pascal and Kepler platforms
> solution for adding the patch will be to create more source tarballs
> (.orig-patch1-$arch.tar.gz) and unpack them, overriding what we unpacked
> first.
Is there any point to uploading 8.0.61, or should we just skip to
8.0.61.2? I haven't been able to do any testing yet, but I am at the
point now where I can try building the SDK samples.
[1] http://docs.nvidia.com/cuda/pdf/CUDA_Toolkit_Release_Notes.pdf
