No, I don’t mean providing infrastructure, platform, or software as a service as Amazon, Google, Salesforce, and numerous others do today, but in terms of generating digital currency – BitCoins.
If you haven’t heard, BitCoin is an experimental peer to peer digital currency currently in virtual circulation. There are about 5.7 million BitCoins in existence according to BitCoinCharts, a website that provides financial and technical data related to the Bitcoin network.
According to BitCoin’s own site, “The competition for coin creation will drive the price of electricity needed for generating a coin close to the value of the coin, so the profit margin won’t be that huge.” They also state that it would take “a year on average to generate 50 coins with a typical PC”.
I wondered whether one could find compute power at a cost low enough to generate BitCoins profitably.
Enter SpotCloud, a Cloud Capacity Clearinghouse and marketplace that enables organizations to buy and sell compute capacity.
Could one or more SpotCloud compute instances (currently starting at $0.001 cent an hour) compute enough hashes per second of BitCoin’s algorithm to produce a net positive financial gain, in effect printing money?
A year of compute is 62,320 hours (to generate 50 BTC) x $0.001 (on SpotCloud) = $61.32. Given BitCoin’s current exchange rate of around $0.80 cents on the BitCoin market x 50 coins = $40. Not a money making proposition. Though it’s unclear what the BitCoin team means by a “typical PC”.
If you’re curious about the arbitrage possibilities, here is a BitCoin generation calculator.
What’s more interesting to me is how quickly people are looking at the Cloud not just to hack passwords or arbitrage possibilities, but for all kinds of possibilities, limited only by our creativity.
If you’re in the mood to donate some BitCoins feel free to send them here:
1ExR3LMmuEptF7THEK823djU6uDftY6az
– Tune The Future –
Stephen
March 31, 2011
Many people have examined whether mining bitcoins using various cloud based computing services would be profitable…in short, not even close. At this stage, you have to invest in the most powerful graphics cards in existence (which cost north of $600 each) to get the most compute power per watt. And even then, you’re likely looking at 6 months or more before you’ll make back the cost of the hardware (and that’s assuming a lot of other people don’t start mining in that time and the exchange rate doesn’t fall).
sophialuden
March 31, 2011
From what I’ve read, it’s actually more cost-effective to get used graphics cards on the cheap than to buy shiny new ones.
freeman
April 3, 2011
Yeah, CPU based mining is just not efficient. You don’t have to spend $600 dollars though, even a ATI 5770 makes an okay miner and sells for about $120.
Note to people unfamiliar with the system: mining is not “useless work”. The system uses “proof of work” (a hard problem solved by the computer) to construct what is essentially a distributed notary system. This notary system is necessary to ensure bitcoins are forged or doubly spent. In exchange for their efforts, miners are rewarded by inflating the currency (50 bitcoin every ~10 minutes right now). The inflation follows a specific schedule and after a certain time will stop. After that, miners will be rewarded by small fees levied on transactions.
It is a very cleverly designed system. Much of the criticism I see of it is based on misunderstandings of how it works.
thinkweis
June 16, 2011
I made a site to help newcomers get started. Check out How to mine bitcoins
vmware Consultant
August 26, 2011
It’s a very useful article and interesting to see how people in Bitcoin do their calculations.
M
January 26, 2012
Shouldn’t the universal use case for a Cloud Currency be as a accounting system for Cloud Services? As in network peering; provide services and immediately clear via the bitcoin entry. Rinse. Repeat.
Large providers can use btc as token for payment while small users could monetize their idle computer resources as a small unit of production in the global cloud. I could let my Acer laptop’s P6000 churn out services and my Road Runner 10/1 Mbps link transfer data and my TiBs of storage host a Tahoe datastore while I’m sleeping, then redeem this stored value in the morning for something else offered by the other users of bitcoin. As it is now, my capital is in lose mode in the use it or lose it game. I’d like to offer-up my already paid for production capacity to other at cut-rate pricing, while converting that value into something that I can actually use at some future time.
So here is what I want: an automated bitcoin billing system that I can use to monetize my own li’l Acer-based Cloud unit. Seems simple enough.