douirc
03-08-2006, 07:55 AM
I'm guessing this has been suggested before, but thought I'd bring it up again as I haven't found a clear and concise technical explanation as to why it wouldn't work.
when you fire up your computer, bios reads the cpu to get some information, including the default cpu multiplier. bios then produces a list of multis available to the user (in my case, 4-10). when i found out that all cpus, regardless of their default multi, have a range from 4 - X, i started to think perhaps bios controls the operating freq of the cpu, and by modding the bios you could easily tell the cpu to run at a higher multi. the problem with this thinking (and this is where i'm looking for some validation) is the bios, or any other component on the board, does not deliver the actual operating freq to the cpu. instead, bios tells the mobo clock what speed to run (default of 200mhz) and that clock freq is sent to the cpu, which itself then multiplies to create its operating freq. so, if bios told the cpu to use a multi of 99, the cpu by default would cap at it's default multi and multipy it by the mobo clock to get it's operating freq, right?
i assume this logic goes the same for the hypertransport. if you did a bios mod to offer soemthing higher than 5, like 10, the chip that controls the HT would know what the max multi is (5) and ignore anything higher than it.
the other reason i originally thought this might be possible is because ABIT's MAXFID (http://www.abit-usa.com/technology/maxfid.php) technology allowed users to overclock the multi by sending the processor a different FID, but i guess that was only plausible on a Barton or TBred chip. I'm guessing sending 11001 to a s939 chip won't force it's multi to 22?!?
anyway, just thought i'd bring it up as we're all looking for the holly grail of unlocked multis, plus, who knows, this could always sparc another new idea :)
when you fire up your computer, bios reads the cpu to get some information, including the default cpu multiplier. bios then produces a list of multis available to the user (in my case, 4-10). when i found out that all cpus, regardless of their default multi, have a range from 4 - X, i started to think perhaps bios controls the operating freq of the cpu, and by modding the bios you could easily tell the cpu to run at a higher multi. the problem with this thinking (and this is where i'm looking for some validation) is the bios, or any other component on the board, does not deliver the actual operating freq to the cpu. instead, bios tells the mobo clock what speed to run (default of 200mhz) and that clock freq is sent to the cpu, which itself then multiplies to create its operating freq. so, if bios told the cpu to use a multi of 99, the cpu by default would cap at it's default multi and multipy it by the mobo clock to get it's operating freq, right?
i assume this logic goes the same for the hypertransport. if you did a bios mod to offer soemthing higher than 5, like 10, the chip that controls the HT would know what the max multi is (5) and ignore anything higher than it.
the other reason i originally thought this might be possible is because ABIT's MAXFID (http://www.abit-usa.com/technology/maxfid.php) technology allowed users to overclock the multi by sending the processor a different FID, but i guess that was only plausible on a Barton or TBred chip. I'm guessing sending 11001 to a s939 chip won't force it's multi to 22?!?
anyway, just thought i'd bring it up as we're all looking for the holly grail of unlocked multis, plus, who knows, this could always sparc another new idea :)