Scientist in France and Singapore are working on processors mimicking the pattern of human neocortex, which seems to calculate probabilities and assigning weights to each of ones, thus push in a kind of exploration mode filtered through a cultural prism.
They intend to use such processors for applications that do not necessarly needs true answers, but where truer than false answers are acceptable, like in graphic GPU.
This will have for effects to decrease the work load of those processors and increase their efficiency compare to the old architecture using pipelines...
Anyhow, any ideas if you think that it has potentials?
For me I was thinking that neuronal computers where the thing that was mimicking the best our brain, but it seems not after all...
So where are we heading?
They intend to use such processors for applications that do not necessarly needs true answers, but where truer than false answers are acceptable, like in graphic GPU.
This will have for effects to decrease the work load of those processors and increase their efficiency compare to the old architecture using pipelines...
Anyhow, any ideas if you think that it has potentials?
For me I was thinking that neuronal computers where the thing that was mimicking the best our brain, but it seems not after all...
So where are we heading?