______ "Dynamic Overclocking" ______

[Blain]

over from the MB forum...

"Dynamic Overclocking" makes perfect sense to me. In the real world I may not need my system running balls to the wall, if it's idle with only a screen saver going.

Gigabyte = CIA & CIA2, MSI = DOT, Others ???
* MB manufacturers design different methods of "DO". They use different "triggers" for their function with various limits.
* Finding real world testimonials on "DO" methods has been extremely hard to pull off.

Can real users of "DO" MBs please post their experiences here?
Please include as much information about your setups as possible.

Thank you very much.

 

[n7]

Originally posted by: Blain
* Finding real world testimonials on "DO" methods has been extremely hard to pull off.

That's cause no real OCer will use auto OCing methods, since they generally result in instability, or tiny baby-sized OCs

/i got nothing for ya, sorry.
No auto OCing for me

 

[GundamSonicZeroX]

Anything that's auto OCing sucks. Plain and simple.

 

[Blain]

Originally posted by: GundamSonicZeroX
Anything that's auto OCing sucks. Plain and simple.

Share with me, your experience.

I'd like to find personal testimonials, because it's too easy to post...
It sux... It's crap... No good... Etc, etc, etc... :Q

 

[myocardia]

Try reading the sticky at the top of this forum (2nd one down), and then, instead of making all of the changes through the BIOS, just change your vcore, vdimm, HTT multiplier, and memory timings in the BIOS. Once you boot to Windows, use ClockGen to raise the HTT/fsb. If you screw up, and go to high, your computer reboots, instead of it stopping working completely, until the BIOS is reset.

And yes, I've actually tried the auto overclocking method once, just for grins, with my Skt. 754 3700, on an MSI Neo Platinum. Like n7 said, it isn't worth messing with.

 

[JAG87]

Dynamic Overclocking is not auto OC. Its basically automatic functions that lower your clockspeed when idling, and put it back up under load. Unfortunately these functions (CIA for gigabyte, NOS for Asus, and DOT for MSI) completely screw up any type of user configured overclock, because they mess with your voltages.

If you want stable rock solid overclocks, stay away.

 

[GundamSonicZeroX]

Originally posted by: Blain

Originally posted by: GundamSonicZeroX
Anything that's auto OCing sucks. Plain and simple.

Share with me, your experience.

I'd like to find personal testimonials, because it's too easy to post...
It sux... It's crap... No good... Etc, etc, etc... :Q

My experiences are all the same with these types of things. It's either OC is unstable and PC craps out and then I have to reseast the CMOS battery, or the OC is something wussy like 15MHz.

 

[GuitarDaddy]

I agree, all the auto OC software I have used sucks and causes instability.

I accomplish DO by using Cool n' Quiet suplemented by RMclock.

I overclock manually in bios with CnQ enabled, then I use RMclock to set min-max multi's and voltages and control when it throttles

Here how I have mine set for 24/7 use with a 3700+ 11x multi

Boots at 10x250=2.5ghz 1.45v, as soon as CPU utililization drops it throttles to 7x250=1.75ghz 1.1v. When CPU utilization goes above 20-25% it jumps to 11x250=2750mhz 1.55v

 

[Blain]

I think I get it...
You OC from the BIOS for max stability, then let the CPU throttle down when idle.

Does Intel's Speedstep function on an OC'd, Speedstep enabled CPU?

 

[GuitarDaddy]

Originally posted by: Blain
I think I get it...
You OC from the BIOS for max stability, then let the CPU throttle down when idle.

Does Intel's Speedstep function on an OC'd, Speedstep enabled CPU?

Not sure as I've never used one. But I'm guessing it would work the same way.
The key IMO is using a utility like RMclock to control the throttling, because if your using CnQ or speedstep alone Windows will control the throttling and it does a poor job IMO. With something like RMclock you have much more control