Author: Robert Hyatt
Date: 17:40:03 02/13/00
Go up one level in this thread
On February 13, 2000 at 16:32:09, Tom Kerrigan wrote: >On February 13, 2000 at 11:05:08, Robert Hyatt wrote: > >>I'd bet that the difference is in picoseconds, not nanoseconds, which doesn't > >Oh, there's something I forgot. > >If you're running at 1GHz, your clock cycle is 1 nanosecond. > >So it's a stupid idea that the switching time would change by nanoseconds (it's >obviously < 1 ns already). > >Moreover, a few picoseconds is pretty significant, compared to 1 ns. > >-Tom Do the math again. If you run at 1ghz (1ns) and drop the temp 40 degrees, and pick up 10 picoseconds, that doesn't make much of a difference. If you picked up 100 picoseconds, you would get a 10% improvement, although you can not do the same for memory. I don't think the operating temperature is the issue with the kryo people. I think it is a function of running Vcc up, running the clock up, and using the freon cooler to keep it from frying. This has been a stock trick for 10 years. Had a company in Huntsville AL, that was using a peltier device to run a 16mhz 386 at 40mhz reliably... back when 386/16's were all you could get. Raising Vcc does the trick, but you have to somehow dump the heat...
This page took 0.02 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.