[cryptography] bitcoin scalability to high transaction rates

Sampo Syreeni decoy at iki.fi
Thu Jun 16 14:35:02 EDT 2011

Since I've been forced to take yet another look into BitCoin and 
algorithmic (high frequency) trading within a short timespan, I began to 
wonder how they would work together. What precisely would happen to 
BitCoin if we had tens to tens of thousands of high frequency traders 
(thousands of transactions per second per trader) within the network?

I haven't been able to come up with any substantive claim about what 
that would lead to, but it at least seems possible that we could hit 
some limits within the distributed log algorithm which could lead to 
significant economies of scale in producing new blocks. Even if we 
actively sparsify the history -- simply knowing about what to start the 
process with could be bandwidth heavy, and keeping up with all of the 
eventually false block chains given that bandwidth could necessitate 
very high end hardware with an unusual, mainframe-kind I/O-cycles 
balance. If that were to happen, it could centralize the verification 
activity to a degree that is amenable to takeover, for simple economic 
reasons. Alternatively, the distributed algorithm could simply become 
choked to a degree via bandwidth constraints (not processing cycles) 
that would lead to enough time inconsistency within the cloud to make it 
almost impossible for it "to prune the bush into a stalk".

Do you think something like this could happen? Is it a viable failure 
scenario? I fear this especially because the incentive payment for block 
creation isn't in any way divisible between those who tried, which means 
that winner takes all, and so that with very high rates of transactions, 
only highly centralized and massive scale processing plants can expect 
to reap a benefit from block processing which is statistically within 
usual financial timescales.

(That is, the reward from contributing processing power to the system 
isn't divisible, unlike BTC's themselves. I can't for example run a low 
power mining operation and expect to get .000001 BTC in a reasonable 
time, whereas somebody with more power statistically will have it in a 
shorter time. That's a clear benefit to scale, because people do not 
have infinite liquidity, and aren't risk neutral.)

(And yeah, this is mostly about economics still. But since BitCoin is a 
cryptographic protocol, this stuff speaks to the threat model and the 
incentive design which is integral to why the protocol is designed the 
way it is. Thus, also to now the protocol should be developed to 
reincentivize better.)
Sampo Syreeni, aka decoy - decoy at iki.fi, http://decoy.iki.fi/front
+358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2

More information about the cryptography mailing list