Houston: Bitcoin has a problem

The recent spike in network difficulty, coupled with the decline in the US Dollar to Bitcoin exchange rate is quickly changing fundamentals of operating a GPU-based mining operation. Whether it’s a single computer with a single graphics card, or multitudes of highly optimized mining rigs, the sun is quickly setting on GPU-based miners.

For instance, according to calculations run on Bitcoinx’s, mining profitability calculator, a freshly deployed mining rig consisting of 3 top-of-the-line Radeon GPU’s will likely take two full years to pay for itself. While the site attempts to compensate for the ever increasing network difficulty, anybody’d guess about how quickly the difficulty level will increae in the future is just that, a guess. A year ago, for instance, no one would have speculated how high the difficulty would be today, and today it’s anyone’s guess as to where the difficulty levels will stand a year from now.

That being said, miners are in a conundrum – ASIC mining devices are the only options that reasonably stand a chance of paying for themselves in a reasonable time frame, but there is an absolute dearth of shipping units. Meanwhile, GPU’s are around and plentiful, but the likelihood that they’ll EVER pay for themselves is slim.

There is the chance that as operators of GPU-based mining farms roll up their operations, they will sell off their Radeon cards at such a steep discount that they look compelling, but even that might not be a winning proposition for miners
who have to pay for their electricity. There exists an exception – if GPU miners move from SHA-256 based coins to those with Scrypt-based proofs of work, there is the definite possibly that they will be able to operate profitably for a longer period, with the flip-side being that if every GPU based Bitcoin mining operation switches to a Scrypt based coin instead, then they will drive difficulty through the roof in in short time.

A glance at the Bitcoin wiki’s Mining Hardware Comparison page lists several ASIC options, but in the course of that glance one will also notice that only a minority of the devices listed are actually shipping to customers. Listed are (in alphabetical order):

ASICMiner is the pre-eminent name on the list, led by an operator who only lets the world know who he is via the pseudonym “friedcat”. Though to date, they have held onto nearly all of their production, utilizing it for themselves and their shareholders, with the exception of the Block Eruptor USB miners that they have produced and shipped. And while the USB device boasts an incridble performance per watt ratio, the absolute performance is less than half of many GPU based solutions.

Avalon ASIC lots #1, #2, and #3 – Note, Avalon was recently forced under the spotlight amid revelations from buyers that they were not shipping the units nearly as quickly as had originally projcected, further confirmed by postings by Avalon co-founder Yifo Guo on Bitcointalk.com’s forums that they were suffering from unspecified yet severe-sounding production issues (to quote Yifo speaking through his Bitcoin Forum handle, BitSyncom, “we fully understand there is problems with orders ( no way?! ) – we can not address these issues IF WE DO NOT SOLVE OUR PRODUCTION ISSUE then we can not DELIVER at all.“)

Butterfly Labs – Butterfly Labs was the first ASIC company to occupy the hot seat of having thousands of customers furious that their products where not nearly as close to production as they had originally thought. Recently, though, Butterfly has been reporting more and more units being shipped out the door, making them them essentially the only real option for customers wishing to obtain ASIC hardware capable of multiple gigahash performance.

The Bitfury Group reports to be working on an ASIC solution, but details are incredibly sparse. Whether they’ll ever have a product to ship would likely be a valid question.

Lastly, KNCMiner lists a couple of mining units under development, but again, with details as sparse as they are, asking whether they have an actualy product in the pipeline is a valid question.

The point is newcomers to the Bitcoin economy have scant options for participating in it. GPU miners are about to become money-losers for anyone who has to pay their power consumption. As GPU farms are taken offline, the Bitcoin network will essentially be run by ASICMiner unless and until Butterfly Labs gets products in the hands of their customers.

The most scary questions is this, though: What if Butterfly Labs comes up dry? Will the Bitcoin network become a single player game? Will users retain the enthusiam they have for it now once a single company has gained 51% control? Afterall, that level of control would normally be considered a successful attack vs the network, in otherwords, not a good thing.

Relevant news