Bitcoin Mining in 2011: GPUs, Scams and the ASIC Arms Race

In early 2011, Bitcoin was still an obscure cryptography experiment. Mining rewards were generous, the wider public had barely heard of it, and much of the community gathered on Bitcointalk to exchange code, mining advice and increasingly questionable business ideas.

I had spent years tinkering with graphics cards, so reading Satoshi Nakamoto’s whitepaper immediately caught my attention. Here was a practical attempt to solve the double-spending problem without relying on a central authority, using proof-of-work to establish a shared history of transactions.

At the same time, Bitcoin mining was moving from ordinary processors to GPUs. Miners were using OpenCL to exploit the parallel processing power of graphics cards—hardware I already understood and could experiment with myself.

The Triple-GPU Frankenstein Rig

I built a Frankenstein mining rig in my living room: three AMD Radeon HD 5870s tethered to a conventional PC by PCIe extension cables, with a high-wattage power supply keeping the whole contraption running around the clock.

It was hot, noisy and far from elegant, but it worked.

I mined through a pool rather than trying to find entire blocks on my own. At its peak, the machine generated regular daily payouts, giving me enough Bitcoin to experiment with more than the mining process itself.

What interested me was whether Bitcoin could actually move value between countries.

I sent coins from my local wallet to a European exchange, transferred them to one of China’s earliest Bitcoin exchanges, converted them into RMB and withdrew the proceeds to an Alipay account.

Exchanges were still involved, so this was not a completely intermediary-free payment system. But I had moved value from Europe to China without making a conventional international bank transfer.

In 2011, that felt remarkable.

The BitScalper Lesson

Early Bitcoin combined genuine technical experimentation with an almost complete absence of consumer protection. The same forums that hosted thoughtful conversations about cryptography and distributed systems were also filled with schemes promising extraordinary returns.

BitScalper presented itself as a Bitcoin arbitrage service. Its operator claimed to exploit price differences between exchanges and share the resulting profits with depositors.

I put some of what I had mined into it.

The returns looked compelling—until the operator disappeared with everyone’s coins. BitScalper had been operating effectively as a Ponzi scheme, using the promise of easy arbitrage profits to attract deposits.

It was an expensive introduction to a pattern that would become familiar throughout the cryptocurrency industry: a plausible technical explanation wrapped around an offer that was simply too good to be true.

Unfortunately, it wasn’t my final costly lesson.

The Butterfly Labs Gamble

As more miners joined the Bitcoin network, GPUs began giving way to application-specific integrated circuits, or ASICs. Unlike general-purpose graphics cards, ASIC miners were designed to do one job: calculate Bitcoin hashes as quickly and efficiently as possible.

I spent several thousand pounds pre-ordering dedicated mining hardware from Butterfly Labs. On paper, the machines promised a dramatic improvement over my GPU rig.

The problem was that they existed on paper for rather a long time.

My order eventually arrived about six months later. The hardware functioned, but the Bitcoin network had not stood still while I was waiting. Mining difficulty had risen so quickly that the machines were already largely uneconomic by the time they reached me.

They were technically capable and commercially useless.

That was one of the defining problems of the early ASIC market. A machine could appear highly profitable when ordered, yet be practically obsolete by the time it arrived. Delivery time mattered almost as much as the hardware itself.

No Convenient Bitcoin Ending

It would make for a satisfyingly neat story to say that early Bitcoin mining transformed my finances, paid off my mortgage or funded my education.

It didn’t.

The Bitcoin itself did not ultimately pay for much. Its more lasting value was the experience of participating in a new technical system before the surrounding industry had matured—or even worked out what it wanted to become.

Years later, I completed the Full-Time MBA at Imperial College Business School, choosing Imperial over an offer from CASS because of its closer relationship with science, engineering and technology. That was a separate chapter rather than the financial consequence of mining Bitcoin.

Looking back, the experience gave me an unusually direct view of both sides of an emerging technology.

I built the hardware and watched mining economics change in real time. I moved value between Europe and China without using a conventional international transfer. I lost coins to a fraudulent investment scheme. I waited six months for specialised hardware that was nearly obsolete when it arrived.

Early Bitcoin contained genuine technical innovation, extraordinary optimism and a great deal of nonsense. Participating in it meant learning to distinguish between them—occasionally only after paying for the lesson.

Opencl

Bitcoin Mining in 2011: GPUs, Scams and the ASIC Arms Race