Slashdot Mirror
Earth Simulator, G5 Cluster Drop In 'Top 500' List
daveschroeder writes "The November Top 500 supercomputer list has been published at SC2004. Topping the charts is IBM and the US Department of Energy's 'BlueGene/L DD2' beta system, at 70.72 TFlops, followed by NASA's 'Columbia' at 51.87.TFlops. For the first time in several publications of this list, Japan's Earth Simulator is no longer in the number one slot, falling to third. Virginia Tech's 'System X' Xserve G5 cluster, while 20% faster than the original cluster that debuted at number 3 last November, has fallen to number 7 due to the new entries, but remains the fastest supercomputer at an academic institution. Here's an excellent cost comparison (Google cache) of the top machines ('System X' is significantly cheaper than anything else in the top 20, not to mention cheaper than many things far below it in performance)."
adminstration and maintenance similar perhaps... but what about power?a few watts per core adds to a lot more heat PLUS the cost of cooling. i think it would be interesting if they printed a FLOP/$ per annum for each of the top 500. the cost of acquisition being spread evenly over the lifetime of the cluster.
Your CPU is not doing anything else, at least do something.
I think it is interesting that 11% of the top 500 are Power architecture, and 64% of the top 500 are intel based systems. Yet 50% of the top 10 are Power architecture and only 20% of the top 10 are intel architecture. Also interesting is that the Power based systems seem to have twice the Mflop/dollar ratio over the intel systems.
That's because x86 is a horrible architecture. On top of that, x86 instructions are translated into microcode before they're executed, so you end up with an unknown (maybe you could ask the folks at Intergraph about it) architecture emulating a crappy architecture in hardware. Better architectures exist (ARM, MIPS, POWER, 68000, PA-RISC, toy architectures used in introductory computer architecture classes, everything else), but Intel won out in the marketplace. You can still get better chips, but you pay more and have less support.
That's why you'd be better off investing in AMD over Intel. AMD hit upon what Intel should've done years ago. The x86-64, for those who don't know, supports x86 binaries as well as its own new architecture. Think of it like an x86 chip with the underlying hardware exposed. If Intel had exposed the hardware that x86 instructions get translated to, they'd have had a clear upgrade path instead of having to dork off x86 out of the blue. AMD embraced and extended x86, and marginalized its future without doing any actual damage to it or x86 users. It's flat out genius.
In the meantime, almost anything performs better than x86, and with less power consumption. It makes those mini-ITX boards look like jokes, because instead of engineering a low powered MIPS board/processor, the VIA folks did another x86. It may have been good from a business point, but it's horrible from an engineering standpoint, and that sums up Intel and x86 fairly nicely.
The cost he quotes for the Blue Gene ($200 million) was the cost of some government contract that included BG/L, ASCI Purple (a huge cluster of POWER5 servers) and some R&D as well.
Recently IBM announced their commercial prices for BG machines (see e.g. theregister.co.uk or news.com.com). Prices start at $1.5 million (1 fully equipped cabinet). Using this price and published linpack figures one arrives at about 2.9 Mflop/s/$, compared to the maximum value of 2.2 Mflop/s/$ he quotes for the best apple system.
Add in the fact that the BG uses much less space and power than a comparable xserve cluster, that it has a faster and lower latency network, and we have a winner.