Why would using 16k servers to run an algorithm be funny? Google runs its algorithms on hundreds of thousands of servers. Do you think if/when Nupic reaches the scale of a human brain, it will run on a single server?
On Mon, Feb 2, 2015 at 3:24 PM, cogmission1 . <cognitionmiss...@gmail.com> wrote: > @Michael Yeah, I was just doing the same kind of invalid comparison to > illustrate my point. Plus I just think using 16k computers is almost the > funniest thing I've ever heard (with regard to an algorithm implementation) > :P > Anway, my fault I totally took the bait :) > > > > On Mon, Feb 2, 2015 at 5:14 PM, David Ragazzi <davidraga...@gmail.com> > wrote: > >> Actually most of these bytes are due to "external" folder which contains >> precompiled static libraries, headers and binaries, for Windows, Linux >> 32/64, and Dawin. >> >> While "external" folder has about 210 MB, the other folders has all >> together less than 4 MB!! Is not fault of HTM algorithms! >> >> On 2 February 2015 at 20:58, Michael Klachko <michaelklac...@gmail.com> >> wrote: >> >>> Lol, please don't shoot! >>> >>> I'm just trying to understand the complexity of HTM algorithms. What >>> would an estimated line count of all HTM algoritms implemented in Nupic - >>> just the algoritms? >>> >>> Btw, your example of 16k servers is wrong on multiple levels. First of >>> all, same code was run on 3 servers two years later with the same results. >>> Second, that code was performing a task whick Grok is not suited for. >>> Finally, even if the task was the same, and Grok was more efficient, why >>> would that mean it needed more code? >>> >> >> >> >> -- >> David Ragazzi >> MSc in Sofware Engineer (University of Liverpool) >> OS Community Commiter at Numenta.org >> -- >> "I think James Connolly, the Irish revolutionary, is right when he says that >> the only prophets are those who make their future. So we're not >> anticipating, we're working for it." >> > > > > -- > *We find it hard to hear what another is saying because of how loudly "who > one is", speaks...* >
