Add a sample with a neural network. The point here would be to have a sample of an FPGA-implemented NN where the whole network, including the weights and functions, is etched into the FPGA fabric.
[ ] This can be a simple NN implementation we provide, we don't need to use a library. Keep perf tips in mind, this needs to be parallelized.
[ ] Add it to Hast.Samples.SampleAssembly and expose it with a SampleRunner like all the other samples there.
[ ] Benchmark the FPGA implementation with the CPU one and add under the Benchmarks docs page.
Add a sample with a neural network. The point here would be to have a sample of an FPGA-implemented NN where the whole network, including the weights and functions, is etched into the FPGA fabric.
Hast.Samples.SampleAssembly
and expose it with a SampleRunner like all the other samples there.Jira issue