GA-EP43-UD3L CPU Termination Voltage questions.

[maestromark]

I have a q8400 this is used in a EP43-UD3L board. I have done extensive testing on stability @ 400fsb. I make a simple change in the CPU Termination Voltage from 1.2 to 1.3 and the needed Vcore to remain stable for 8 hours of Linpack testing @ 90% memory usage drops by .025 to .050 which is very nice. From what I understand the raising slightly of the CPU Vtt helps clear up some of the signal noise when going to high FSB rates. Can that explain the reduction of needed Vcore to remain absolutely stable and why?

Thanks!

[yuu]

Re: GA-EP43-UD3L CPU Termination Voltage questions.

Probably it has something to do with you using more MCH, and Vtt needing to be in close range with MCH voltage or the CPu cores themselves being connected by Vtt.

I set 1.00V to Pll, Vtt and Mch and it didn't demand any more votlage on 415,2 FSb that is my exact limit with 1.25Vcore (at 415.8, 416.4 and 417.0 it fails reporting a succes having finished ~200 out of 2000 loops).

Also at 500Fsb the DRam Fine delays would change with some vtt cpu dram combinations, rendering the system unstable, but not because it needed more vtt.

[maestromark]

Re: GA-EP43-UD3L CPU Termination Voltage questions.

Well currently the CPU Termination Voltage is Currently 1.30 and its normal setting is 1.20
The current MCH Core Voltage is 1.20 from a normal setting of 1.10 volts. I would appreciate a further discussion on the interdependencies of these settings and how they affect the CPU stability in relation to a high FSB of 400. The MCH Core had to be raised from 1.10 to 1.20 to get stability of the CPU once I pushed the FSB above 333 as a guide line rule in my stability testing. I could get away with a Vtt of 1.20 and no adjustment there. I did not understand another way to get a FSB of 400 on the other hand. Also the EP43-UD3L does not have any type of LLC adjustment in the bios as a option as well. And my top FSB limit appears to be around 417 FSB and that is either CPU or RAM limited not exactly sure which is holding me back or could be the top end of this CPU.

Thanks

[yuu]

Re: GA-EP43-UD3L CPU Termination Voltage questions.

CPU ref and MCH ref are moving up with VTT. 0.76 -> 0.82. maybe you'll get the same result just by upping to 0.82.

[Lsdmeasap]

Re: GA-EP43-UD3L CPU Termination Voltage questions.

Yes, all the ref voltages change when you raise Vtt or ram voltages, they appear to stay the same but this is only a percent of the other said voltages.

See the MCH GTL ref guide in my signature for more info on what voltages change what and how that works