Slashdot Mirror
Asus Releases Desktop-Sized Supercomputer
angry tapir writes "Asustek has unveiled its first supercomputer, the desktop computer-sized ESC 1000, which uses Nvidia graphics processors to attain speeds up to 1.1 teraflops. Asus's ESC 1000 comes with a 3.33GHz Intel LGA1366 Xeon W3580 microprocessor designed for servers, along with 960 graphics processing cores from Nvidia inside three Tesla c1060 Computing Processors and one Quadro FX5800."
More importantly can this actually run Crysis 2? Probably not.
While I wouldn't choose to do my scientific computing on Windows, I know some people do, and those Tesla cards (which are providing the bulk of the processing power) really don't care which OS you're running.
[b.belong('us') for b in bases if b.owner() == 'you']
"Supercomputer" might mean cluster, a big node (to go in a cluster), or big-iron mainframe.
It's not a cluster, and it's not much of a mainframe, but it has a helluva lot of FLOPS for a single node. To me, it looks similar to the nodes that went into Roadrunner's TriBlades - 2 Opterons (as general purpose processors) plus 4 PowerXCell 8i (for heavyweight vector processing), and a total of 16G memory. But I'm not an expert.
Still, I bet that if you could hook 3240 of them together, you would have a strong Top500 contender.
Supercomputer is a computer that is one of the most powerful computers available at a given time. Therefore referring top500 list is very valid when determining what is supercomputer and what isn't. Top500 list can very well be used in determining whether we have a supercomputer or not. If the modern computer isn't faster (at least in certain specific tasks) than the lowest performing computer on the list I wouldn't consider it being a supercomputer. I don't understand the need to dilute supercomputer word to include cheap hacks like this, there are valid names for these such as minicomputer. What do we call the best performing computer? superdupercomputer?
San somebody who has actually worked with such machines enlighten me about its performance on tasks that are not floating point intensive? Our simulations mainly push many,many objects around, with relatively little, or no floating point math in them.
Do such machines still make sense, or are we better off with a bunch of general purpose CPUs clustered together? How do they compare to Suns Niagara cpus that have umpteen hardware threads in them ?
Complexity is a measure of our ignorance...