Originally Posted by dellandre
Ok, I'm stuck on step 7. I see my "Minimal FID" is at 6.0x at the managment tab, but I dont see anything with "Maximal FID." I also don't even see a apply button. What am I missing?
Your image is a little too small there mate
, but I'm guessing you're following the guide on the first page; the program has been updated since then. Here's an update I wrote for it, hope it helps:
Originally Posted by Mr. K6
hmm... I think there needs to be a a few things updated for 1.6v, here goes:
First off, go to the general tab and select whatever functions you want on the bottom, but make sure that you select "restore CPU defaults on exit" in case something messes up, then you can just exit
. Then, under the management tab, select use p-state transition. In the white box right below that there should be two (you'll probably have to create at least one). The first should be State No. 0, with an FID of 5x (or whatever your lowest is) and a voltage of 0.998v (or w/e it is stock on your lap top). This is the 800MHz power saving setting for the new P-Ms. The other should be State No. 1, with your max FID (mine is 15x) and your stock full voltage (mine is 1.356v).
Make sure that your A/C adaptor is plugged in, then go down to AC profile and select "maximal" from the drop down box. Get Prime 95 started and calculating (Large FFTs work best for a test). Also make sure that you select "Use OS-based load management" (same as run HLT etc.). From here go back to your P state transitions box and click on the max setting (the one with your highest multiplier) and modify the voltage going to it. I took it down 3 steps at a time until I got to about 1.15 and then I went down one at a time from there. Make sure you hit apply on the bottom of the management tab after each voltage change. You can also use CPU-z to vertify the voltage change. If you're pin-modded you may need to stop sooner because your CPU will require more juice. Keep lowering it until you get an error, then raise it and prime test. Anything higher than 8 hours means you're rock solid. I got an instant error at 1.054, but I run 1.1v just for the extra stability (10hrs+ stable).
Once that has been tested, go back to the management tab and under AC profile select "minimal". This will use your 800MHz CPU (or whatever your lowest multi is) setting. The odd case here is that now you will load at this speed, which will never happen in "real-world," the CPU is only at 800MHz when it's idle. In the same way drop your voltage down (remember to hit apply). I think almost every CPU can do .75v, and most can do .700v, but test and test again. After that let prime95 run for 3 hours or so, which is more than enough because the CPU will never be loaded at this state.
Ta-da, you're done. Just make sure you go to management and select "Automatic Management" under your AC Profile. Also under you General tab select "state minimized" and "run automatically at start up," but make sure you keep "restore defaults upon exit" just in case something ever glitches on your. Have fun and rock on
Great job attix! That's a damn fine chip you have there, you might try shoving it in desktop using that ASUS board and seeing how far you can take it. That's what I'm going to do with mine once I need a new desktop: put a pin-modded 1.6 in this lappy and overclock the hell out of my 2.0GHz
ivar, I think you totally missed what I said, so I'll try clarifying it. I said:
Originally Posted by Mr. K6
What Intel specifies is far beyond the actually requirements of the processor. That's why we can undervolt our Dothans
. I would almost bet that that ULV chip could do lower
voltage than yours at that speed (stably, of course).
Notice the addition
. Nowhere did I say proportionally or comparitively, I just said that it would be lower. That said, and let's just say all the chips are capped at .700v, I'd still bet that the ULV can do 1.1Ghz @ .700v, which is still lower
than what these standard P-M's can do, which was the point of my comment. However, even if the ULV's are capped at .700v, that's an artifical roadblock on their potential; who knows, maybe they can do it at .550v or even lower