Right - had another little fiddle around.
Decided to try cranking the FSB straight up to 200, and run the memory at 83% to run at it's stock PC2700 speed.
Seems perfectly stable (although need more time to tell) - and the temperatures have not moved more than the usual margin of error... (ie not at all). So this Barton is more than capable of sitting happily at 2.2GHz (3200+) speeds it would seem.
Now the question is (which with a couple more days running CPDN I'll be able to answer myself): how much performance is lost by running the memory asynchronously like this? Do I lose more than I gain by ramping the CPU up 20%?
Decided to try cranking the FSB straight up to 200, and run the memory at 83% to run at it's stock PC2700 speed.
Seems perfectly stable (although need more time to tell) - and the temperatures have not moved more than the usual margin of error... (ie not at all). So this Barton is more than capable of sitting happily at 2.2GHz (3200+) speeds it would seem.
Now the question is (which with a couple more days running CPDN I'll be able to answer myself): how much performance is lost by running the memory asynchronously like this? Do I lose more than I gain by ramping the CPU up 20%?


I'm running off the MCP ethernet adapter now btw, but the drivers are just those from the cd-rom and then windows-updated... as I'm only using the one port at the moment, I'll give it a go from the 3com one with the MCP port disabled and see what difference it makes...
Comment