It would be interesting to see if this ends up delivering us a commodity processor dedicated to Double Precision math... that would be a game changer in HPC no doubt... bitcoin seems to require that level of precision... I wonder if you can implement it with the mixed precision method... hmm... https://en.bitcoin.it/wiki/PHP_developer_intro#Precision
On Apr 16, 2013, at 10:09 PM, Tom Metro wrote: > Bitcoin has been all over the news this past week. I think the main > trigger was some significant price swings on the exchange rate. Then > there were follow-on stories talking about how mining activities are > using $150,000 in electricity every day. > > Even if you don't care about alternative currencies, you may find the > hardware used to enable "mining" for coins to be of interest. Here's a > story talking about said hardware: > > http://gizmodo.com/5994446/digital-drills-the-monster-machines-that-mine-bitcoin > > The basic idea with Bitcoin is that you have a computer do some > intensive computational work, specifically compute hashes: > > http://en.wikipedia.org/wiki/Bitcoin#Bitcoin_mining > > The mining process or proof-of-work process involves scanning for a > value that when hashed with SHA-256, the hash begins with a number of > zero bits. The average work required is exponential in the number of > zero bits required, but can always be verified by executing a single > hash. > > And your reward for doing the work is you get a coin, which can be > exchanged for goods and services, or for US dollars on an exchange. > > A clever part of the design is that it self-adjusts for the ever > increasing capabilities of the available hardware: > > To compensate for increasing hardware speed and varying interest in > running nodes over time, the proof-of-work difficulty is determined by > a moving average targeting an average number of blocks per hour. If > they're generated too fast, the difficulty increases. > > And this has led to: > > Today, bitcoin mining is a competitive field. An arms race has been > observed through the various hashing technologies that are used to > mine bitcoins and confirm transactions: High-end GPUs (Graphical > Processing Units) common in many gaming computers, FPGAs (Field > Programmable Gate Arrays) and ASICs (Application-specific integrated > circuits) all have been used. The newest addition, ASICS, are built > into specialized servers that can cost nearly $3000 USD a unit. > > Also listen to: > http://www.grc.com/sn/past/2011.htm#287 > > for Steve Gibson's explanation of how the algorithm works. > > In any case, the competition and the cost of power is driving miners to > FPGAs and ASICs. Pretty amazing that this made up currency scheme has > been profitable enough to justify the development of ASICs. > > I have to assume that with this degree of profit motive, there must be > criminal organizations using zombie bot networks to do mining. When the > cost of the hardware and electricity is free, you clearly have an advantage. > > The Gizmodo article has some interesting pictures showing racks of > open-frame computers, from CPU-era to GPU-era to FPGAs and finally > ASICs. They describe on ASIC solution (quoting Gizmodo): > > Meet the Avalon ASIC. ... Inside its unassuming grey case is an army > of specialty chips that promise 65 gigahashes per second. This is > $6,800 equipment...that, properly utilized, stands to be worth much > more. > [...] > And while Avalon started the revolution, others are not far behind. > Companies like ASICMiner, Butterfly Labs, and bASIC all offer similar > systems--some that boast even more power--but none of those have > managed to ship. > [...] > The Butterfly Labs BitForce Mini Rig SC, a 1,500 GHz/s machine. It's > about 30 times more powerful than then current Avalon ASICs, or would > be if it ever exists... > [...] > Bitcoin has always favored early adopters, bold ones especially. > Because mining difficulty increases to compensate for increased power, > all advantages are temporary. But it doesn't increase dynamically. > Instead it hops up once every 14 days, so advantages can be leveraged. > The first Avalon ASIC to be put to use paid for itself in a mere nine > days, but every additional unit follows will have more and more ground > to make up. > > The treatment of the tech in this article is a bit superficial. It would > have been nice if they described in more detail what was depicted in > each photo. But probably a decent overview of the hardware, if you only > have a casual interest in Bitcoins. > > A concluding thought: > > Ultimately, ASIC devices are the last great innovation in Bitcoin > mining; once you've specialized down to the chipset, there's nowhere > left to turn for a 100-fold computing power increase. And in that way, > we're seeing the beginning of the end of the gold rush, just as > Bitcoin fever reaches a fever-pitch. > > That's kind of a defeatists view, and doesn't seem to acknowledge Moores > law. Over time the ASICs will shrink and you'll see more hashing engines > per chip. > > How about Quantium computing? > > More likely, the next innovation will come not from hardware but from > mathematics - finding an algorithmically more efficient way of arriving > at the same result. > > -Tom > > _______________________________________________ > Hardwarehacking mailing list > [email protected] > http://lists.blu.org/mailman/listinfo/hardwarehacking
_______________________________________________ Hardwarehacking mailing list [email protected] http://lists.blu.org/mailman/listinfo/hardwarehacking
