Is this the first Pi supercomputer ?

no, it's not a supercomputer, it's a cluster.

Supercomputers are rated in FLOPS.  You won't see a rating for this.

A desktop PC with two 8GB dram sticks has as much total memory.

RasPi only lets you program the ARM11 directly.  There's also a bunch of video and graphics processors, along with a DSP (or so I've read), none of which are documented.  I wonder how much performance you could get from a single RasPi if you could program those processors as well?

I've heard Eben say the GPU has 24 GFLOPS of general purpose computing.  This cluster had 64 Pi's so that would be 1.5 TFLOPS I believe.  Of course the trouble is there's no way to tap into that yet.  I asked Eben about it back in May at Maker Faire and he commented they would like to open it up but the interface won't be OpenCL.  I'm not sure what the current plan is.

Even the claimed 24 GFLOPS (single precision only), times 64 Pi's (assuming perfect performance scaling across 100Mbit ethernet implemented over USB 2.0) giving 1.5 TFLOPS in the ideal case, assuming it could be opened up for the user to program, is not competitive with a single modern GPGPU graphics card with 3.2 TFLOPS single precision.

http://www.amd.com/us/press-releases/Pages/amd-powerful-server-graphics-2012aug27.aspx

Virginia Polytechnic University in the U.S. made, the cheapest supercomputer to date (in 2000) using a few handfuls of powermac g4's. I think it was the third or fourth fastest supercomputer at the time, at a small fraction of the cost

With reference to the Virginia Polytechnic system,the lead designer says " With dual 64-bit, 2.3GHz G5 processors in each of its 1,100 Xserve G5 units, the new cluster operates at 12.25 teraflops".

Is there some confusion betwen clusters and supercomputers ?

(See http://www.apple.com/science/profiles/vatech2/)

> Is there some confusion betwen clusters and supercomputers ?

Wikipedia defines a supercomputer as:

"A supercomputer is a computer at the frontline of current processing capacity, particularly speed of calculation." http://en.wikipedia.org/wiki/Supercomputer

The Virginia Polytechnic system was ranked #3 on the list of the world's

top 500 supercomputers in 2003, #7 in 2004, #14 in 2005 with a rating of

12.25 teraflops, and #280 in 2008, so there's no ambiguity there.

In contrast, the Pi cluster has no FLOPS rating, no high-speed

interconnect, and a very small ram, and is clearly not at the frontline of

current processing capacity.   So no ambiguity there either.

The benchmark that was run to calculate pi is designed to show

the speedup from running multiple nodes in parallel, but they only

showed it running on two nodes, and deleted the timing information.

The RPi has been measured at about 40MFLOPS under Raspbian,

so a two-node RPi cluster is clearly not a supercomputer.

For teaching purposes, with the budget they used, they actually could

have made an instructional cluster that also had impressive performance,

but not with RPi's.