Top500 Supercomputers

[icaci]

The funny thing is that FLOPS means floating point operations (per second) and SHA256 computations (the core Bitcoin hashing algorithm) take exactly 0 (zero) floating point ops as they consist of integer and bitwise ops only. Besides floating point operations are usually slower than integer and logical ones.

Bottom line is - measuring Bitcoin performance in FLOPS is wrong!

[1715602432]

1715602432

[MoonShadow]

The funny thing is that FLOPS means floating point operations (per second) and SHA256 computations (the core Bitcoin hashing algorithm) take exactly 0 (zero) floating point ops as they consist of integer and bitwise ops only. Besides floating point operations are usually slower than integer and logical ones.

Bottom line is - measuring Bitcoin performance in FLOPS is wrong!

That's true, but nearly every supercomputer in the top list are optimized for floating point operations.  That metric is simply an estimate of relative performance, so that the user can, at a glance, compare apples to oranges.

[proudhon]

The funny thing is that FLOPS means floating point operations (per second) and SHA256 computations (the core Bitcoin hashing algorithm) take exactly 0 (zero) floating point ops as they consist of integer and bitwise ops only. Besides floating point operations are usually slower than integer and logical ones.

Bottom line is - measuring Bitcoin performance in FLOPS is wrong!

Sure, measuring bitcoin hashing performance in flops is wrong, but what we want to know is what the flops of the computers in the bitcoin network would be if they were doing something that required floating point operations?  Maybe there's just no way to know, but that's the question, I think.

[marcus_of_augustus]

The funny thing is that FLOPS means floating point operations (per second) and SHA256 computations (the core Bitcoin hashing algorithm) take exactly 0 (zero) floating point ops as they consist of integer and bitwise ops only. Besides floating point operations are usually slower than integer and logical ones.

Bottom line is - measuring Bitcoin performance in FLOPS is wrong!

Sure, measuring bitcoin performance hashing performance in flops is wrong, but what we want to know is what the flops of the computers in the bitcoin network would be if they were doing something that required floating point operations?  Maybe there's just no way to know, but that's the question, I think.

Well, I think you've got that backwards. The question is, in terms of "strength" of the network it is more important to know what the hash power of the competition could likely be, rather knowing the flop power of our network .... i.e. they are not going to attack us with flops but hashes.

So the flops-equivalent of BTC network number should properly be converted into a hash-equivalent of a possible attacker ... but I think people are mostly too lazy to bother doing that so just do it backwards and think forwards, if you know what I mean.

[proudhon]

The funny thing is that FLOPS means floating point operations (per second) and SHA256 computations (the core Bitcoin hashing algorithm) take exactly 0 (zero) floating point ops as they consist of integer and bitwise ops only. Besides floating point operations are usually slower than integer and logical ones.

Bottom line is - measuring Bitcoin performance in FLOPS is wrong!

Sure, measuring bitcoin performance hashing performance in flops is wrong, but what we want to know is what the flops of the computers in the bitcoin network would be if they were doing something that required floating point operations?  Maybe there's just no way to know, but that's the question, I think.

Well, I think you've got that backwards. The question is, in terms of "strength" of the network it is more important to know what the hash power of the competition could likely be, rather knowing the flop power of our network .... i.e. they are not going to attack us with flops but hashes.

So the flops-equivalent of BTC network number should properly be converted into a hash-equivalent of a possible attacker ... but I think people are mostly too lazy to bother doing that so just do it backwards and think forwards, if you know what I mean.

Oh, I think you're right that the more useful question is the question of how much hashing power other computer systems have.  But, some people just want to know how many flops the bitcoin network has for this reason:

[icaci]

http://www.top500.org/

#1 Tianhe-1A ; Rpeak = 4701 Tflops
http://www.top500.org/system/performance/10587

#2 Cray "Jaguar";  Rpeak = 2331 Tflops
http://www.top500.org/system/performance/10184

Combined total of Top500 supercomputers is 43673 Tflops
http://www.top500.org/lists/2010/11/performance_development

#0 Bitcoin Grid GPU/CPU ; Rpeak = untested, estimate 15106 Tflops
http://www.bitcoinwatch.com/

Tianhe-1A hashing power, very accurately IMO has been estimated as 850 Ghps. Assuming the rest of top 500 supercomputers have the same TFLOPS/Ghps ratio than bitcoin needs to hit 8Thps to dwarf all top 500 supercomputers combined. We are more than halfway there already. Another difficulty increase or two and 50% attack with supercomputers would require all those babies in top 500 list combined. Moreover while they configure them to do something bitcoin network would double once more.

Tianhe-1A is a mixed architecture: 14366 CPUs + 7168 GPUs. A rough estimate of hashing power is 14366 * Xeon X5670 (8 Mhps) + 7168 * nVidia M2050 (100 Mhps) = 831 Gh/s. Xeon X5670 peak performance is rated as 2.93 GHz * 4 DP * 6 cores = 70.32 DP Gflops (DP == double-precision floating point as used to measure performance in Top500). nVidia states that M2050 achieves 515.2 DP Gflops. Thus for the CPU the Gflops/Mhps ratio is 8.8 and for the GPU it is 5.2. The combined CPU+GPU ratio is 5.7. For a single HD5870 GPU this ratio is 1.5 (544 Gflops and 375 Mhps).

There are few mixed CPU/GPU computers in Top500. Amongst the top 10 are only Tinahe-1A, Nebulae and TSUBAME. Roadrunner is using Cell processors which are similar to CPUs in performance. There are only 10 systems using nVidia technology at the Top500 list from November 2010. All other supercomputers are only CPU based thus the combined Gflops/Mhps ratio for the whole Top500 group would be around 8. At 43.6 Tflops combined this leads to an estimate of about 5.5 Thps. We are currently at 3.5 Thps.

As most miners are running GPU and not CPU miners, I think the estimated Tflops rate of the Bitcoin network is a very large overestimate: 42895 Tflops / 3377 Ghps = 12.7 is a ridiculous ratio. It's most likely that the ratio should be around 2 or 3, depending on the portion of remaining CPU miners.

[marcus_of_augustus]

Are we there yet?

http://www.bitcoinwatch.com/

Network Hashrate Gigahashs/s    4768.66
Network Hashrate TeraFLOP/s    60562 (equivalent)