The obsession most mining people have with bitcoin is really fascinating. You people tend to overestimate the importance of bitcoin relative to the GPGPU and parallel processing as a whole. This leads to some absurd conspiracy theories like AMD research driven by bitcoin or bitcoin mining as driving force in new GPU sales. Reality is rather different. Computing resources thrown at bitcoin are very likely even less than those in SL3 unlocking. Assuming the average miner has 500MH/s at his disposal, the whole mining society would be comprised of less than 30.000 users. The amount of GPUs thrown at BTC mining is really insignificant doesn't matter how important it is to you. This is even more valid as far as FPGAs are concerned.
Unlike what you may think, there is something called "computational finance" and it has some embarassingly parallel applications, for example Monte Carlo simulations used for financial planning. The investment JPMorgan made for hardware would bring them much more profit if used for those tasks as opposed to bitcoin mining.
Also unlike what you think, corporate customers and universities are still way bigger market as far as GPGPU/FPGA niche is concerned as compared to hobbyist and semi-hobbyist uses like bitcoin mining and SL3 unlocking.
Bottom line: the universe does not revolve around you, sorry.
All valid points, but I think you're underestimating the impact of Bitcoin mining at least in some niches. E.g. Vladimir buying all 5970's in the UK
(he's running a big mining contract service and I think he was serious with that comment).
why don't they pay people to run computational finance crap on their GPUs? I mean if you pay more than people mine in BTC then those terahashes would be all theirs
Because of the myriad well-known disadvantages of grid computing? (Secrecy, protection against fake work, protection against tampering, communication overhead, need for a suitable platform for distributing the work, being at the mercy of end users who may quit at any point or not be convinced at all to join in sufficient quantities...). Bitcoin mining is very unique in its suitability for monetized grid computing.
They're doing fine without your Terahashes, thank you very much (and of course, they don't need literally hashes).
why don't they pay people to run computational finance stuff on their GPUs? I mean if you pay more than people mine in BTC then those terahashes would be all theirs
Great idea for a business. If someone wrote the components to make this happen I'm sure there would be very lucrative applications.
One which I (and surely others) have had for years. But like I said, for most application there are severe limitations for this kind of deployment, which I guess is why nobody managed to pull off something successful yet.