Now I'm virtually all FPGAs and am desperate to get more hashrate (I'm a committed, but very small, miner), but FPGAs are expensive and only compete with GPUs on power consumption - in my experience (best hash per £ GPUs, so 5770, 5830, 5850, 6950 all with special tweaks; and a combination of Ztex's 1.15d and 1.15x FPGAs, so all Spartan-6 units).

I lost out on bASIC Tom even though I considered my due diligence pretty good; he had a record of delivering and didn't send off my risk analyst radar like BFL did. And whilst Avalon *have* delivered ASICs, I can't take days off work glued to my computer, propped up on amphetamines and repeatedly hitting ⌘R on the slim chance that I'll get onto the order webpage - and even then needing the luck to be able to actually *pay* and get a guaranteed build slot. I fear now is far too late anyway and Avalon have already sold their entire short production run, so it's 'start the process again'. I wouldn't be surprised if BFL go the same way, since there are too many rumours of chips not making the cut, and like Avalon they would only have a very limited number of discrete chips to package and test. With all the pre-orders, I'm sure I've missed the boat

Sure, I have the option to spend another £5k on 25 of Stefan's wonderful little boards and build another cool looking rig. It will be ultra-efficient and use little power. But, like the current one, it only has around 5 GH/sec. The ASIC devices are looking like around 50 GH/sec for something around £1k to (probably) £4k. I had a paid-for ASIC order in along with the other early adopters, but unluckily my order was with Tom and not Avalon. So if the ASIC boat has gone, with terahashes of power in various states of assembly at BFL and Avalon (et al) already earmarked for customers, buying as many FPGAs as possible *now* doesn't sound like it'll ever pay off  And that's even if I could get the FPGAs in the first place. Perhaps I should sell my existing FPGA rig for Bitcoins and simply invest (buy and hold only)...