Fabian Tschopp

@dagamayank Some of the values can be autotuned (such as WPTM, WPTN), others are defined by the convolution settings (such as v_p, v_s, v_d). However the autotuner can't store the...

@dagamayank Have you made any progress on this or is something too complicated?

@dagamayank I understand. I will work on the project this weekend and hopefully have some improvements until monday. One interesting thing I found is that I'm better off targeting TLP...

@dagamayank You can remove the layers after the 1st convolution in the prototxt file, or start with any other convolution as long as you have the input data defined &...

@dagamayank LibDNN is now available as a standalone library: https://github.com/naibaf7/libdnn

@zazd Yes it uses a local-memory and register-level GEMM. It is similar to cuDNN, you can read up more here: https://arxiv.org/pdf/1410.0759.pdf

@bhack @gstoner Good news for the RX 480: Performance issues and thermal envelope crashes have been completely fixed since Linux kernel 4.8 AMDGPU drivers. It is now possible to use...

@bhack For the external library I did not port them yet... Quite busy with a new project at the moment regarding sparse RNN's. :) Let me know if you need...

@bhack Yeah last week, Codeplay's CEO contacted me regarding some stuff in OpenCL TensorFlow. If he expresses interest as well, I will definitely re-focus more on the libdnn standalone. But...

No, this version of OpenCL Caffe is not suitable to FPGAs due to how the OpenCL kernels are compiled and optimized. You could try other projects targeting FPGA: https://github.com/BenBBear/caffe-fpga-opencl