Bitcoin network hashrate PetaFLOPS 195.66 vs Worlds supercomp 16.32 petaflops/s

[istar]

MANNHEIM, Germany; BERKELEY, Calif.; and KNOXVILLE, Tenn.—For the first time since November 2009, a United States supercomputer sits atop the TOP500 list of the world’s top supercomputers. Named Sequoia, the IBM BlueGene/Q system installed at the Department of Energy’s Lawrence Livermore National Laboratory achieved an impressive 16.32 petaflop/s on the Linpack benchmark using 1,572,864 cores.

http://top500.org/

Bitcoin Network Hashrate
PetaFLOPS   195.66

How can the Bitcoin network hasgrate be more than 10 times, one of the worlds most powerful supercomputers.
I cant see how some enthusiasts in just 3 years have surpassed worlds nr 1 supercomputer by 10 times!

Is that really correct or are they messured with different methods?

[Nemesis]

MANNHEIM, Germany; BERKELEY, Calif.; and KNOXVILLE, Tenn.—For the first time since November 2009, a United States supercomputer sits atop the TOP500 list of the world’s top supercomputers. Named Sequoia, the IBM BlueGene/Q system installed at the Department of Energy’s Lawrence Livermore National Laboratory achieved an impressive 16.32 petaflop/s on the Linpack benchmark using 1,572,864 cores.

http://top500.org/

Bitcoin Network Hashrate
PetaFLOPS   195.66

How can the Bitcoin network hasgrate be more than 10 times, one of the worlds most powerful supercomputers.
I cant see how some enthusiasts in just 3 years have surpassed worlds nr 1 supercomputer by 10 times!

Is that really correct or are they messured with different methods?

do you really need to ask?

I guess BFL SC minirig will beat worlds supercomputer?

Hell no.

[iain]

MANNHEIM, Germany; BERKELEY, Calif.; and KNOXVILLE, Tenn.—For the first time since November 2009, a United States supercomputer sits atop the TOP500 list of the world’s top supercomputers. Named Sequoia, the IBM BlueGene/Q system installed at the Department of Energy’s Lawrence Livermore National Laboratory achieved an impressive 16.32 petaflop/s on the Linpack benchmark using 1,572,864 cores.

http://top500.org/

Bitcoin Network Hashrate
PetaFLOPS   195.66

How can the Bitcoin network hasgrate be more than 10 times, one of the worlds most powerful supercomputers.
I cant see how some enthusiasts in just 3 years have surpassed worlds nr 1 supercomputer by 10 times!

Is that really correct or are they messured with different methods?

They're measuring wildly different kinds of computation - real number (floating point) calculations for weather simulations, weapons simulations, etc etc (for which "flops" [floating point operations per second] makes sense) vs. endlessly repeated integer-based hashing (for which "flops" has not the slightest relevance).

STATEMENT BANDIED ABOUT IN LOTS OF PLACES (on this forum and elsewhere): "No way can anyone ever mount a 51% attack, because the Bitcoin network is far more powerful than all the world's supercomputers combined." (or variant wordings with greater or lesser degrees of hyperbole)

TRANSLATION TO REALITY: It's a few thousand graphics cards. A mere footnote in the budget of any bank or government that wanted to outpace it. We do need to worry about how to stand up to 51% attacks, 90% attacks... the full gamut. Let's start thinking and worrying now!

(My own thoughts: https://en.bitcoin.it/wiki/Proof_of_blockchain_fair_sharing. And yes, I'll say it before anyone else does, these thoughts are still hopelessly unfinished - barely more than started - but I've published them in unfinished form because I want to get as many people thinking about this as possible!)

[Gabi]

And we have another noob comparing cpu with gpu, comparing bitcoin mining with things wich require tons of memory and super fast connections between cores...

The next comparison will be "omg a ferrari is better than a tir for hauling things: it goes much faster than the tir!"?

[sturle]

MANNHEIM, Germany; BERKELEY, Calif.; and KNOXVILLE, Tenn.—For the first time since November 2009, a United States supercomputer sits atop the TOP500 list of the world’s top supercomputers. Named Sequoia, the IBM BlueGene/Q system installed at the Department of Energy’s Lawrence Livermore National Laboratory achieved an impressive 16.32 petaflop/s on the Linpack benchmark using 1,572,864 cores.

http://top500.org/

Bitcoin Network Hashrate
PetaFLOPS   195.66

How can the Bitcoin network hasgrate be more than 10 times, one of the worlds most powerful supercomputers.
I cant see how some enthusiasts in just 3 years have surpassed worlds nr 1 supercomputer by 10 times!

Is that really correct or are they messured with different methods?

This is 100% pure ox manure.  The Bitcoin network is exactly 0 FLOPS.  Zero floating point operations are involved.

Do not compare integer operations with floating point operations.  It is two different worlds.

[unclemantis]

Seems like we are wasting a lot of computer cycles just to stitch blocks together We should focus on solving world peace.

[kjj]

And we have another noob comparing cpu with gpu, comparing bitcoin mining with things wich require tons of memory and super fast connections between cores...

The next comparison will be "omg a ferrari is better than a tir for hauling things: it goes much faster than the tir!"?

Check that again, the table states figures are a corrected value to give a flops representation of non floating point operations. In terms of hashes per second bitcoin is well ahead of any other system, good luck using that as a benchmark though, I doubt you'll find many takers.

Correction is the root of all evil.

There are three people standing in a line.  Would you weigh them and then use a table to give the "corrected value" of the number of people?  Or would you count them?  Counting is easy, the FLOP count for bitcoin is still zero.  Zero FLOPS, zero FLOPD, zero FLOPY.

If you still insist on "correcting" the value, explain your correction scheme, and justify why you picked it instead of the other options available to you.*  Do you convert by considering how many integer operations it would take to emulate the corresponding floating point operation in the same chip?  Do you look at the idle FPU and calculate how much work it could be doing if it was doing work?

Do you take work distribution and networking into account?  Why or why not?  "Official" FLOPS scores are found by using a brutal test suite that requires an extensive amount of coordination and control, while bitcoin is looking for a magic needle with an army of needle factories, each of which is allowed to run with almost zero coordination.  The bitcoin network is simply not suitable for LINPACK.

Bitcoin is an amazing system, and viewed as a whole, the network uses a staggering amount of computing power.  But all of that power is being applied to a problem that was chosen very carefully to allow the network to exist and function.  We are awesome at what we do, the best in the world by leaps and bounds, and not even the largest HPC clusters can match what we do.  But we can't match what they do either.

* Yes, I am fully aware that no one in this thread is doing the correction.  My point is that the correction is unphysical, even when your source is trying really hard to do it well.

[ArticMine]

MANNHEIM, Germany; BERKELEY, Calif.; and KNOXVILLE, Tenn.—For the first time since November 2009, a United States supercomputer sits atop the TOP500 list of the world’s top supercomputers. Named Sequoia, the IBM BlueGene/Q system installed at the Department of Energy’s Lawrence Livermore National Laboratory achieved an impressive 16.32 petaflop/s on the Linpack benchmark using 1,572,864 cores.

http://top500.org/

Bitcoin Network Hashrate
PetaFLOPS   195.66

How can the Bitcoin network hasgrate be more than 10 times, one of the worlds most powerful supercomputers.
I cant see how some enthusiasts in just 3 years have surpassed worlds nr 1 supercomputer by 10 times!

Is that really correct or are they messured with different methods?

They're measuring wildly different kinds of computation - real number (floating point) calculations for weather simulations, weapons simulations, etc etc (for which "flops" [floating point operations per second] makes sense) vs. endlessly repeated integer-based hashing (for which "flops" has not the slightest relevance).

STATEMENT BANDIED ABOUT IN LOTS OF PLACES (on this forum and elsewhere): "No way can anyone ever mount a 51% attack, because the Bitcoin network is far more powerful than all the world's supercomputers combined." (or variant wordings with greater or lesser degrees of hyperbole)

TRANSLATION TO REALITY: It's a few thousand graphics cards. A mere footnote in the budget of any bank or government that wanted to outpace it. We do need to worry about how to stand up to 51% attacks, 90% attacks... the full gamut. Let's start thinking and worrying now!

(My own thoughts: https://en.bitcoin.it/wiki/Proof_of_blockchain_fair_sharing. And yes, I'll say it before anyone else does, these thoughts are still hopelessly unfinished - barely more than started - but I've published them in unfinished form because I want to get as many people thinking about this as possible!)

Gavin Andresen has raised this issue in http://gavintech.blogspot.ca/2012/05/neutralizing-51-attack.html

[Trader Steve]

Seems like we are wasting a lot of computer cycles just to stitch blocks together We should focus on solving world peace.

By wrestling control of the monetary system out of the hands of central planners you effectively strip away their ability to tax. No taxes means no funding for endless wars. Sounds like a great way to bring about world peace to me.

[istar]

Seems like we are wasting a lot of computer cycles just to stitch blocks together We should focus on solving world peace.

By wrestling control of the monetary system out of the hands of central planners you effectively strip away their ability to tax. No taxes means no funding for endless wars. Sounds like a great way to bring about world peace to me.

Taxes, what do you need it for when you can print money?

[fivemileshigh]

The ability to fund wars via unfounded future liabilities and inflation is arguably greater than that supplied by outright taxation, so bitcoin is still a major step forward.

I guess a more relevant question is, should those who control some of the most powerful computers on that list use them against the bitcoin network, what kind of damage would we be looking at?

[cypherdoc]

Seems like we are wasting a lot of computer cycles just to stitch blocks together We should focus on solving world peace.

By wrestling control of the monetary system out of the hands of central planners you effectively strip away their ability to tax. No taxes means no funding for endless wars. Sounds like a great way to bring about world peace to me.

Taxes, what do you need it for when you can print money?

taxes?  oh yes, they need it, they want it, and they're coming for it.

[aq]

Seems like we are wasting a lot of computer cycles just to stitch blocks together We should focus on solving world peace.

By wrestling control of the monetary system out of the hands of central planners you effectively strip away their ability to tax. No taxes means no funding for endless wars. Sounds like a great way to bring about world peace to me.

Taxes, what do you need it for when you can print money?

taxes?  oh yes, they need it, they want it, and they're coming for it.

Actually that would be an interesting concept. If the IRS would not collect any tax, but, based on some estimations about the overall economy size, would just "print" the tax money. This way everyone would "pay" the same percentage as a result of the implicit currency devaluation.

[cypherdoc]

Seems like we are wasting a lot of computer cycles just to stitch blocks together We should focus on solving world peace.

By wrestling control of the monetary system out of the hands of central planners you effectively strip away their ability to tax. No taxes means no funding for endless wars. Sounds like a great way to bring about world peace to me.

Taxes, what do you need it for when you can print money?

taxes?  oh yes, they need it, they want it, and they're coming for it.

Actually that would be an interesting concept. If the IRS would not collect any tax, but, based on some estimations about the overall economy size, would just "print" the tax money. This way everyone would "pay" the same percentage as a result of the implicit currency devaluation.

you've done it!  why hasn't anyone thought of this before?!  the path to perpetual prosperity and it won't cost any of us a thing!

edit:  that's the problem, printing was never designed to benefit ALL of us.