[HH] Bitcoin mining hardware

[Tom Metro]

We're going astray of hardware hacking, so this will be my only
contribution to this branch of the thread.

Here are some quotes from a Paul Krugman column (an economist that
writes for the NY Times). He doesn't seem to understand the technology
of Bitcoins very well, and makes some factual errors. (He describes them
as anonymous and untraceable, which is wrong. My understanding is they
can be anonymous, but they are always traceable, in that all
transactions are rigorously logged.)

He does, however, make a good point about the significance of currency
being backed by precious metals.

http://www.nytimes.com/2013/04/15/opinion/krugman-the-antisocial-network.html

  The biggest declared investors in bitcoins are the Winklevoss
  brothers...and they make claims for the digital product similar to
  those made by goldbugs for their favorite metal. "We have elected,"
  declared Tyler Winklevoss recently, "to put our money and faith in a
  mathematical framework that is free of politics and human error."

  The similarity to goldbug rhetoric isn't a coincidence, since goldbugs
  and bitcoin enthusiasts -- bitbugs? -- tend to share both libertarian
  politics and the belief that governments are vastly abusing their
  power to print money.

  At the same time, it's very peculiar, since bitcoins are in a sense
  the ultimate fiat currency, with a value conjured out of thin air.
  Gold's value comes in part because it has nonmonetary uses, such as
  filling teeth and making jewelry; paper currencies have value because
  they're backed by the power of the state, which defines them as legal
  tender and accepts them as payment for taxes. Bitcoins, however,
  derive their value, if any, purely from self-fulfilling prophecy, the
  belief that other people will accept them as payment.
  [...]
  Goldbugs and bitbugs alike seem to long for a pristine monetary
  standard, untouched by human frailty. But that's an impossible dream.
  Money is, as Paul Samuelson once declared, a "social contrivance," not
  something that stands outside society. Even when people relied on gold
  and silver coins, what made those coins useful wasn't the precious
  metals they contained, it was the expectation that other people would
  accept them as payment.

Even if the current reality of Bitcoins is that it is dominated by
"bitbugs" playing games of speculation, as hardware hackers we should
take interest in the hardware used to perform the mining (which could
ave more practical applications, as Kurt pointed out), and appreciate
the beauty of the Bitcoin cryptographic algorithm.


Greg London wrote:
> from the next paragraph:
> 
> ;>Once a predetermined number of coins have entered
> ;>circulation, the incentive can transition entirely
> ;>to transaction fees and be completely inflation free.
> 
> This makes no mathematical sense. The value of these "coins"
> must neccessarily go down as computational power goes up.

You're missing a key point. The algorithm adjusts for computational
advances. Every 14 days it corrects the amount of work that needs to be
done to earn a coin to maintain a fixed rate of so many coins per hour.

http://en.wikipedia.org/wiki/Bitcoin#Bitcoin_mining
  To compensate for increasing hardware speed and varying interest in
  running nodes over time, the proof-of-work difficulty is determined by
  a moving average targeting an average number of blocks per hour. If
  they're generated too fast, the difficulty increases.


The paragraph you quoted is referring to this:

http://en.wikipedia.org/wiki/Bitcoin#Distribution
  Currently, 25 bitcoins are generated every 10 minutes. This will be
  halved to 12.5 BTC within the year 2017 and halved continuously every
  4 years after until a hard-limit of 21 million bitcoins is reached
  within the year 2140. As of March 2013 over 10.5 million of the
  total 21 million BTC had been created; the current total number
  created is available online. In November 2012, half of the total
  supply was generated, and by end of 2016, three-quarters will have
  been generated. By 2140, all bitcoins will have been generated with
  the last one consisting of fractional parts.


> When that PDF tries to go into the "why" of bitcoins,
> it invariably starts rambling about not wanting to
> trust a third party. "Not trusting" taken to a point
> becomes "paranoia".

My take on the "why" is pretty simple: today we have cash for any
transaction we wish to be anonymous. If in the future most transactions
happen online, then it may be desirable to have a cash equivalent.

Whether that's a good thing largely comes down to your philosophy and
politics.


Even if Bitcoin doesn't become a general currency, it still might have
uses. We've seen a bunch of schemes to reduce spam through some form of
payment, either in terms of a micro-transaction or performing some
computational work. What if your mail server required the "payment" of a
fraction of a Bitcoin before it would accept mail from a
non-white-listed sender? The sender could chose to buy or mine the
Bitcoins. The recipient could save those coins for its outbound mail or
simply discard them.

The advantage to Bitcoins over other schemes is that it is a relatively
proven algorithm, it automatically compensates for computational
increases, and there is an open market for them so senders can choose to
buy.

(Of course spammers also conveniently control large bot networks, which
could also be used to mine Bitcoins. I see further down in the Wikipedia
article it does report evidence of that having happened.)


> If I understood the math, it wouldn't take much
> to crank out these "coins" using the FPGA fabric
> in the zedboard.

No, such hardware has already been outclassed. It'll still work, but
it'll take so long as to be impractical.


> (3) people looking to launder money.

Yes, quite likely.


> I think I'll just steer clear of it all.

Probably wise, but we can still appreciate the tech.

 -Tom