Training data sets

Publicly available datasets

Dataset name	Source
Tensorflow	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/tensorflow.zip
AMD APP SDK	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/amdappsdk.zip
BLAS	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/blas.zip
PolybenchGPU	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/polybenchgpu.zip
Rodinia	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/rodinia.zip
SHOC	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/shoc.zip
Linux kernel	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/linuxkernel.zip
OpenCV	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/opencv.zip
Eigen (synthetic)	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/eigen-synth.zip
GEMM synthetic	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/gemm-synth.zip
Stencil synthetic	http://spclstorage.inf.ethz.ch/projects/ncc/datasets/stencil-synth.zip

Datasets available for download upon request

In addition to the publicly available datasets, you may want to train on data sets which can be downloaded upon request:

Dataset name	Source
Parboil	http://impact.crhc.illinois.edu/Parboil/parboil_download_page.aspx
NAS	http://aces.snu.ac.kr/software/snu-npb/
Nvidia samples	Bundled with the CUDA toolkit