I came across this article this morning and was thinking about other applications of offloading computations to the GPU. I previously looked at the PS3, and the possibility of an Erlang cluster of PS3s; the problem was the lack of RAM. That killed it for the PS3, to run enough green processes, Erlang’s memory requirements would have saturated ran and paging would kill performance. On a high end GPU there is a bunch of RAM. I’ll look into CUDA and potential Erlang ports to CUDA. I guess another possibility is dispatching semi complex jobs using Erlang, that way you inherit the error processing capabilities of Erlang and then dispatching jobs written in CUDA to the GPUs (Or maybe I’m just over engineering a solution to a problem I don’t have).
A programmer on Edwards’ staff suggested trying to run the models on graphics processing units (GPUs). (GPUs or graphics cards are specialized chips that run inside PCs to display 2D and 3D graphics. They tend to contain hundreds of floating point processors that are good at handling mathematically intensive and parallel processes such as Monte Carlo simulations.) The programmer ran a proof of concept in March 2008 using the cash flow generation part of the algorithm and showed a dramatic increase in performance. That programmer now runs the team of technologists that work on the bond pricing system.
Show me more… »
Tags: Bonds,CUDA,erlang,GPU,parallel computing,PS3,Wall St
Categories: Technology
No Comments »
