I was catching up on my reading and came across an interesting article in the September 2005 IEEE Computer Society "Computer" magazine. The gist of the particular article was that the graphics processors on many high-end PC video cards are overlooked for applications requiring heavy number-crunching. What caught my eye was the chart that illustrated that the Nvidia G70 graphics processor now performs at about 170 GFlops! (A dual-core Pentium 4 running at 3 GHz, by contrast, will do about 20 GFlops). Granted, this is 32-bit vector floating point arithmetic, but the raw numbers are pretty stunning. It simply doesn't seem that long ago that Neil Lincoln was telling me how wonderful his liquid nitrogen-cooled high-density CMOS vector supercomputer was going to be in that it would be the first to break the 10 Gflop mark. (This was the ill-fated ETA Systems one and only ETA-10 supercomputer). Now we have 170 GFlops on the desktop to draw pictures! Times certainly do change. Does anyone know what became of very few ETA-10's that were built? Museum or landfill? Cheers, Chuck

Chuck> I was catching up on my reading and came across an interesting Chuck> article in the September 2005 IEEE Computer Society "Computer" Chuck> magazine. The gist of the particular article was that the Chuck> graphics processors on many high-end PC video cards are Chuck> overlooked for applications requiring heavy number-crunching. Chuck> What caught my eye was the chart that illustrated that the Chuck> Nvidia G70 graphics processor now performs at about 170 Chuck> GFlops! (A dual-core Pentium 4 running at 3 GHz, by contrast, Chuck> will do about 20 GFlops). 20 GFlops with a 3 GHz processor? How can that be -- that would require 3 FP functional units per core, each issuing one instruction per clock cycle. Did you mean 2 GFlops? Chuck> Granted, this is 32-bit vector floating point arithmetic, but Chuck> the raw numbers are pretty stunning. Sure is. There may be as many as three processors tucked away in your PC that are more powerful (though more narrowly focused) than the main processor. The graphics engine is one -- the other two are the digital signal processing engines in the disk read channel and (if you have one) the Gigabit Ethernet interface. There's a group somewhere (can't remember the name or URL) working on developing parallel processing algorithms that run on your graphics card -- essentially treating it as a vector coprocessor for your PC. paul

On 9/23/2005, Fred Cisin wrote: It wasn't just for Macs, was it? Didn't Don Lancaster have some sort of scheme for running math on any Postscript-capable printer? Cheers, Chuck

Inevitably, even inferior technologies catch up at some point. :) *scnr* -- ---------------------------------- personal: http://www.armory.com/~spectre/ -- Cameron Kaiser, Floodgap Systems Ltd * So. Calif., USA * ckaiser at floodgap.com -- It's the car, right? Chicks dig the car. -- "Batman Forever" ---------------

There was also various programs for the C64, mostly demos, that used the 1541 as a co-processor. This was largely hampered by the 1541's paltry RAM and the slow IEC bus, though. -- ---------------------------------- personal: http://www.armory.com/~spectre/ -- Cameron Kaiser, Floodgap Systems Ltd * So. Calif., USA * ckaiser at floodgap.com -- You do not have a soul. You are a soul. You have a body. -- C.S. Lewis -----

On 9/23/2005, Paul Koning wrote: Could be smoke-and-mirror type of benchmarking, but here's the quote: "However, said University of Virginia assistant professor David Luebke, a 3-GHz, dual-core Pentium 4 Extreme Edition's arithmetic units will theoretically run as much as 24.6 Gflops." Admittedly, there are some weasel words there like "theoretically" and "as much as", but it might also be that University of Virginia assistant professor David Luebke moved a decimal point somewhere. Cheers, Chuck