There's also http://arxiv.org/pdf/1505.02142.pdf in which key HTM concepts were adapted to the Heidelberg Neuromorphic Computing Platform.
On Sun, Sep 27, 2015 at 12:42 AM, Pascal Weinberger < [email protected]> wrote: > Hey! > > > I think this may be of interest to you: > > http://www.science20.com/physics_foundations/blog/numenta_and_ibm_to_build_biologically_inspired_intelligent_machines-155769 > > There has also been some GPU implementations of HTM > https://github.com/222464/ContinuousHTMGPU > > Best, > > Pascal > > ____________________________ > > BE THE CHANGE YOU WANT TO SEE IN THE WORLD ... > > > On 27 Sep 2015, at 09:13, Sam <[email protected]> wrote: > > Dear all: > > There has been a lot of work on hardware acceleration of machine learning > algorithms with FPGA, ASIC or GPU, especially for neural networks. I wonder > if it makes sense, or if there is any prior work, to build custom hardware > to accelerate HTM/NUPIC, in order to achieve real-time performance in an > embedded environment? > > Thanks! > > Sam Gu > >
