Forgot your password?
typodupeerror

Comment: Re:A little thin on tech detail (Score 1) 43

by Macchendra (#45105605) Attached to: Qualcomm to Build Neuro-Inspired Chips
A the math for a neural network accelerator chip is indistinguishable from the math for graphics. It's all the same multiplying vectors and matrices. And most of the work is done up front in training the neural network, the results of which can be distributed to computers with processors of lesser power. If qualcomm is going to get a couple of teraflops on a single core then power to them.

Comment: Re:A little thin on tech detail (Score 1) 43

by Macchendra (#45100333) Attached to: Qualcomm to Build Neuro-Inspired Chips
Not a helpful answer. There is no information anywhere on the web on whether these are continuous or discrete neural networks. As far as putting it on the chip, if it is a discrete neural network, then there is no advantage over a cuda enabled neural network running on an nvidia tesla. It is just Malibu Stacy with a new hat.

Comment: Re:Wrong objective. (Score 3, Informative) 115

by Macchendra (#44946517) Attached to: Mozilla Plan Seeks To Debug Scientific Code
It is easier to find bugs in code where all of the objects, variables, methods, etc. are named according to their actual purpose. It is easier for other researchers to integrate their own ideas if the code is self documenting. It is easier to integrate with other software if the interfaces are cleanly defined. It is easier to verify the results of intermediate steps if there is proper encapsulation. Also, proper encapsulation reduces the chances of unintended side-effects when data is modified outside of scope.

What this country needs is a good five dollar plasma weapon.

Working...