Bitcoin mining seems like a good match for data centers. But is it a viable business model?
In the surface, mining for crypto-currencies seems like an ideal application for data centers. The technology requires lots of power and significant cooling, yet places little demand on expensive data center technologies such as high availability, replication and backup. A web search will find dozens of sites that are promoting cloud-based currency mining, but drilling down into the technology paints a confusing picture.
The best known, and most valuable of the crypto-currencies, bitcoin, has had its ups and downs, with a peak value of more than $1,000 crashing quickly to the $500 range and continuing a low decline to where it seems to have stabilized in the mid to high $200 range at the time of writing. And with such a volatile currency, it is unsurprising that companies set up to take advantage of bitcoin mining have been equally unstable.
The brief history of crypto-currency mining is filled with a litany of complaints to regulatory authorities, as everything from basic mining hardware to large-scale cloud mining entities have failed to live up to the claims made by the vendors. The fledgling crypto-currency sector has been plagued with hardware announcements that took pre-orders for products that never saw the light of day, and vendors declaring bankruptcy after losing tens of millions of dollars of investor and customer money.
Few things highlight the duality of the currency mining business better than events of January 2015. First, currency mining darling BitFury, fueled by additional venture capital investment in 2014, bought immersion-cooling company Allied Controls to help it build better, more efficient data centers focused on currency mining. And then well-known bitcoin mining company CoinTerra declared bankruptcy, having managed to hit almost all the highpoints of bitcoin failure, including late or non-delivery of hardware, failure to refund monies owed, and not paying its colocation data center bills. Having failed at almost every aspect of the currency mining business, it left the business with somewhere in the range of $10m to $50m in liabilities.
Even today, anyone looking into the state of crypto-currency mining will find a surprisingly large number of dead links and suspended web pages, even from reference websites that are actively updated. We decided to investigate two different approaches to currency mining: actual mining, with equipment bought from hardware vendors, and the use of third-party cloud-based mining services. The complexity involved in determining where consumers should spend their money is significant, and not for the faint of heart.