We cache models once they’ve been built to save time. I’m wondering if there’s something else going on that is sometimes making it slow. The thing that’s strange to me with the build, is that sometimes nengo_ocl is almost as fast as nengo (128 dims, one point), and sometimes nengo is almost as slow (1024 dims, one point). It’s just benchmark_circconv.py but you wrote a thing to plot the build times, too, right? Can you post your benchmarking script?ĮDIT: Never mind. (running the benchmark code provided at BLAS setup for Numpy page, mentioned by previously.)ĭotted two (1000,1000) matrices in 1138.7 msĮigendecomp of (1500,1500) matrix in 23.233 sĭotted two (1000,1000) matrices in 21.4 msĮigendecomp of (1500,1500) matrix in 8.060 sīenchmark results of nengo simulation with BLAS enhanced numpy: Now numpy utilizes all the cores flawlessly. Someday I will list out the steps I took to set up the stable system, after I test nengo-ocl really works.īig thanks to your suggestion, BLAS setup really works! (Yeah I’m relatively new to Linux and all these tools.) Originally I thought this is an easy job, but it took me entire two days of work and reinstalling entire Ubuntu three times to get to current state. I had run some examples for CUDA toolkit and pyopencl, I think they worked.Īs to nengo-ocl, I’ve try to run the example files found here: īut without instruction, I actually don’t know how to make it put out some useful information. NVIDIA binary driver - version 367.57 from nvidia-367Ĭuda compilation tools, release 8.0, V8.0.44 Excitingly, we (my lab) recently set up a new PC just for Nengo simulation.ĬPU: Intel® Core™ i7-6700K CPU 4.00GHz × 8
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |