Universal Rule

Book Reviews

Thursday, October 23, 2003

The Apple G5 dual supercluster achieves 8.16 teraflop/s? -- Comment()
Slashdot reported that Big Mac Benchmark Drops to 7.4 TFlops. This was expected, and rather high still.

However, there is a preliminary report (in pdf, dated October 22, 2003) of the bechmark results in supercomputing compiled by Jack Dongarra. See page 53 for the highly parallel benchmark for supercomputers. On this list the G5 cluster at Virginia is performing at 8164 gigaflop/s, or 8.16 teraflop/s. Thus, the machine would be at position 4 of the top-500 list. And there might still be room for improvement in the benchmark speed.

Here is a listing of the top-5 systems from the draft report:

System Nr of procs R_max R_peak

Earth Simulator 5120 35860 40960

ASCI Q AlphaServer 8160 13880 20480

HP RX2600 Itanium 2 1936 8633 11616

Apple G5 dual 2112 8164 16896

ASCI Q AlphaServer 4096 7727 10240

Note that the Apple G5 dual result are not for the whole system of 2200 processors. All the speeds are in gigaflop/s (10^9 floating point operations per second).

-- Comment()
Testing Apple's iSight: "The iSight is a well-designed webcam with features which will be incorporated into other companies' webcams in the near future, no doubt. Pity then, that Apple didn't pull out all of the stops to make it more than just a very nice looking and expensive webcam; its automatic functions could have been a lot better." (The Star via MyAppleMenu) [MyAppleMenu]

-- Comment()
Human Genome story: "John Sundman, a man who had three children with rare, horrible diseases, discovered that the human genome was being mapped for profit, with the intention of rendering the results into a proprietary storehouse to be sold to the highest bidder. That's why he's taken a compelling interest in the human genome project, a successful effort to map the genome and turn the data over to the public domain." [Boing Boing Blog]

Preliminary speed of the G5 cluster: 7.41 teraflop/s -- Comment()
I updated my story about the G5 cluster at Virginia with preliminary benchmark results. The R_max figure of 7.41 teraflop/s would put this supercluster to position 4 in the top-500 list of supercomputers. Not bad at all.