Optimize Reshape - Githubissues

waylonflinn / weblas

GPU Powered BLAS for Browsers :gem:

MIT License

702 stars 43 forks source link

Open waylonflinn opened 8 years ago

waylonflinn commented 8 years ago

The GLSL shader for Tensor.reshape is very complex. It also currently accounts for ~15% of the execution time of the NN demo.

waylonflinn commented 8 years ago

Some progress on this. Created an optimized (but un-benchmarked) version for the no padding (multiple of four on input and output) case.