On Sat, Jun 18, 2011 at 11:28 AM, Warren Smith <[email protected]> wrote: > > Oh, I guess you'll say that if there were any faster implementation of the > key > algorithms like 10X faster on some better hardware or with better compiler > or > something, the market would have already found it?? Vomit.) >
Yes. The market has already been quite aggressively optimizing the SHA-256 kernel, which is the heart of the proof-of-work. The Bitcoin network's capacity today is heavily dominated by leading-edge GPGPU cards which do vector-parallel processing of hundreds of block candidates simultaneously. I and others are tinkering around with designing special-purpose architectures for SHA-256 on FPGAs, but we are finding it extremely difficult to match the cost-performance of the existing GPU-based systems. And, the moment a better technology comes along (such as heavily-pipelined SHA-256 ASICs in a leading-edge silicon process), whoever develops it would be much more rational to sell that technology widely to the Bitcoin mining community, rather than trying to use it to directly attack and implode Bitcoin. (What would be the point of that?) Barring algorithmic breakthroughs (like a discovery of a way to find SHA-256 collisions that is much faster than brute search), it seems extremely implausible to me that any one entity would ever amass enough advanced technology to out-compute the entire Bitcoin mining community - or would even want to if they could. -Mike -- Full name: Michael Patrick Frank Email addr.: [email protected] (pers. email) Snail mail: 820 Hillcrest Ave., Quincy, FL, 32351-1618 Phone/voicemail: (413) 842-6670 (main number, uses Google Voice) Webpage URL: http://www.facebook.com/M.P.Frank (pers. profile)
---- Election-Methods mailing list - see http://electorama.com/em for list info
