Bitcoin Forum
January 19, 2018, 06:03:50 AM
 Home Help Search Donate Login Register
 Pages: [1]
 Author Topic: How many transactions per second can the bitcoin network sustain?  (Read 4269 times)
stakhanov
Full Member

Offline

Activity: 175

 February 28, 2011, 07:40:01 PM

It seems that there is a limited number of transactions that can be included in a block. I'd like to have an idea of how much the bitcoin network can scale up?
1516341830
Hero Member

Offline

Posts: 1516341830

Ignore
 1516341830

1516341830
 Report to moderator
1516341830
Hero Member

Offline

Posts: 1516341830

Ignore
 1516341830

1516341830
 Report to moderator
1516341830
Hero Member

Offline

Posts: 1516341830

Ignore
 1516341830

1516341830
 Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
SmokeTooMuch
Legendary

Offline

Activity: 872

 February 28, 2011, 07:56:53 PM

well you can't make a per second statement, since, in theory, their could be thousands of blocks in one second.

and because all of the transactions in one block get processed at once, it's hard to say how many transactions per second there are.

I guess you are looking for something like this:
"How many transaction fit in one block ?" and "Is there a maximum of blocks per second in the network ?"

Date Registered: 2009-12-10 | I'm using GPG, pm me for my public key. | Bitcoin on Reddit: https://www.reddit.com/r/btc
barbarousrelic
Hero Member

Offline

Activity: 675

 February 28, 2011, 08:29:29 PM

well you can't make a per second statement, since, in theory, their could be thousands of blocks in one second.

and because all of the transactions in one block get processed at once, it's hard to say how many transactions per second there are.

I guess you are looking for something like this:
"How many transaction fit in one block ?" and "Is there a maximum of blocks per second in the network ?"
Isn't there an average number if blocks generated per unit time that will roughly remain constant over long enough time periods? (as difficulty adjusts to compensate for changes in network computing power)

"How many transactions can fit in one block ?"

should get us close to the answer we're looking for.

edit: according to the FAQ

Quote
The average rate of block creation in the whole system is 6 / hour.

So the rate is 6*(max transactions per block) per hour.

Do not waste your time debating whether Bitcoin can work. It does work.

"Early adopters will profit" is not a sufficient condition to classify something as a pyramid or Ponzi scheme. If it was, Apple and Microsoft stock are Ponzi schemes.

There is no such thing as "market manipulation." There is only buying and selling.
Mike Hearn
Legendary

Offline

Activity: 1526

 February 28, 2011, 09:03:04 PM

The network can probably sustain around 80 tps right now, assuming some significant fraction of nodes are running on single core machines. The bottleneck is the ECDSA verification of inbound transactions.
jgarzik
Legendary

Offline

Activity: 1512

 February 28, 2011, 09:25:32 PM

You'll hit maximum block size, before ECDSA becomes a bottleneck.

Jeff Garzik, Bloq CEO, former bitcoin core dev team; opinions are my own.
Visit bloq.com / metronome.io
Donations / tip jar: 1BrufViLKnSWtuWGkryPsKsxonV2NQ7Tcj
stakhanov
Full Member

Offline

Activity: 175

 February 28, 2011, 10:11:57 PM

The network can probably sustain around 80 tps right now, assuming some significant fraction of nodes are running on single core machines. The bottleneck is the ECDSA verification of inbound transactions.

Wouldn't this be reached very quickly if the bitcoin system starts having any kind of success? How could we improve that rate?
Gavin Andresen
Legendary

Offline

Activity: 1652

Chief Scientist

 February 28, 2011, 10:48:33 PM

80 transactions per second is 6.9 million transactions per day.  That is a LOT.

PayPal was doing about 2 million transactions per day in 2008; if there are three times as many bitcoin transactions as PayPal transactions in a few years that would be a very good problem to have, and I'm confident "we" would figure out how to deal with lots of transactions.

How often do you get the chance to work on a potentially world-changing project?
stakhanov
Full Member

Offline

Activity: 175

 February 28, 2011, 10:56:21 PM

80 transactions per second is 6.9 million transactions per day.  That is a LOT.

PayPal was doing about 2 million transactions per day in 2008; if there are three times as many bitcoin transactions as PayPal transactions in a few years that would be a very good problem to have, and I'm confident "we" would figure out how to deal with lots of transactions.

Thanks for putting things in perspective. I agree, if bitcoins start generating more transactions than paypal, it will already be a huge success, and we will all be millionaires
randomguy7
Hero Member

Offline

Activity: 528

 February 28, 2011, 11:04:17 PM

How big would the blocks get when we hit that limit? Won't this first become a problem?
jgarzik
Legendary

Offline

Activity: 1512

 February 28, 2011, 11:08:57 PM

See this thread for discussion of how block size limit caps transaction rate.

The block size limit caps bitcoin at 463 transactions/minute, or ~7 transactions per second.

Jeff Garzik, Bloq CEO, former bitcoin core dev team; opinions are my own.
Visit bloq.com / metronome.io
Donations / tip jar: 1BrufViLKnSWtuWGkryPsKsxonV2NQ7Tcj
Mike Hearn
Legendary

Offline

Activity: 1526

 March 01, 2011, 12:17:25 PM

Yeah, there's an artificial limit on block size which can be removed/put off by tweaking a constant and having people upgrade. It's there for valid reasons though.

Increasing CPU capacity of the network is a thornier problem, it can be increased a bit with a software upgrade but not just cranked up arbitrarily like block size can. ECDSA batch verification might change this dramatically.

BitCoin can scale, but indeed I am wrong about where it grinds to a halt today. Right now block size is the limit, as jgarzik says.
Dixie
Newbie

Offline

Activity: 5

 March 02, 2011, 04:43:45 AM

That a good idea.
 Pages: [1]