Hi Chris, Certainly I agree with your assessment that we won't be seeing affordable dev boards based on this chip soon if ever, but I don't think rolling your own FPGA solution is likely to be the quickest or best solution either.
I did a lot of work with FPGAs in the past and would not bet on them to achieve better performance than GPUs. Certainly FPGA solutions will not scale as easily as GPU solutions. Plus there is already a start on porting the CLA to openCL, http://github.com/Jontte/CortiCL. If I wasn't happy after that, I'd see about getting one of these http://apt.cs.manchester.ac.uk/projects/SpiNNaker/hardware/. On Sun, Aug 10, 2014 at 10:48 AM, Chris Albertson <[email protected]> wrote: > On Sun, Aug 10, 2014 at 2:58 AM, cogmission1 . > <[email protected]> wrote: >> Is it worth the $20 to read the article? Is the chip something that could >> host a CLA? > > There is no way you are going to be able to get one of these chips any > time soon. If you want a hardware implementation of CLA your best > option is to buy a FPGA development board. They cost about $100. The > FPGA would likely be about as fast as the chip but has the advantage > that you can change the programming quickly. > > The technology is changing so fast that I don't think most people > would even want a custom chip as it would be obsolete quickly. The > programmable hardware is probably the way to go if software is to > slow. > > -- > > Chris Albertson > Redondo Beach, California > > _______________________________________________ > nupic mailing list > [email protected] > http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org -- James Bridgewater, PhD Arizona State University 480-227-9592 _______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
