Wednesday, December 24, 2008

Parallel Processors and Biological compilers

Hi ,

Merry Xmas and prosperous New Year , everybody .

Leslie , Thanks for the CUDA heads-up . (See http://www.mpronline.com "Parallel Processing with CUDA" sent by Leslie)

I was not aware that commercially available parallel processing and programming systems were so advanced .

All we need is a compiler built into the the assembly and we have a neuronal network .

Note the strong analogy with biological systems . (Not surprising)
Biological systems have threads (axons , dendrites , synapses , glials , inter-glial fluids , nano-tubes , etc)

But which parts play the role of compilers ? Are they always necessary ? Are there many different architectures , evolved for a specific purpose ?

Note the argument in http://andreswhy.blogspot.com "Unpacking and Packing"
Compiling can be seen as packing information , and running the program as unpacking .

The Compiler can be seen as a Beth(1) component . Directed probabilities .
Chaperones spring to mind for unpacking . As expected , it is simpler than packing . (Programming equivalents would be Java-applets , etc)

The Packer can be seen as the Evolutionary Process , and the Compiler as a part of it .

But the Compiler is a specific hardware component .
What is it known as ? It must have been seen and described , just not recognized .

Any help would be gratefully received .

Andre

No comments: