Dell 14z undervolting

with all the multipliers on 1V it hasnt made much diff to the heat produced..has it made a huge impact in urs?

 

@krap101- can u take a ss of the advanced CPU settings page ?

 

\

There ya go

 

A word of caution: I tried this on Windows 7 32-bit last night and when I got down to the level at which it gave me a BSOD, it was unrecoverable. As soon as Windows would boot up and RMClock started, the computer would shut off.

I had to boot up in safemode to disable it. There doesn't seem to be an actual uninstall program - very poor, IMO, so I can't remove it from the registry and reinstall using safer settings.

 

Use something like CCleaner to remove it from the registry.

 

OK, I recently made the switch to using Windows 7 full time, and so I've only just re-done the undervolting. Thought I would share with you the results:

First image here is the readings after a 10 minute stress test without undervolting. My notebook is running a P8600 - it's the Inspiron 1545 and not the 14z, but I'm just showing you these to demonstrate that temperature savings can be achieved and you can undervolt some processors below 1 volt:



OK, so now here is the shot of the same screens, but after I had undervolted the 9x multiplier to 0.9875. I get BSOD when I go to anything below 0.975, and I prefer to be safe and keep it two "notches" above the voltage I know is definitely unstable.



So a temperature drop of 5 degrees Celsius, but more importantly to me, the fan never switched to full speed (and really loud) mode; whereas it did without undervolting.

 

Found out that the program folder contains a .reg file to wipe the settings clean and start fresh. It's called RMClock_WipeOut.reg

 

Yup it does, I still prefer to use ccleaner anyway, as the other method leaves the keys in place but blanks them out I believe.

 

Well after all was said and done, 0.95 V seems to be my minimum safe level.

Temps before were 72, 74 abd 74 on the 9400. After were 62, 63 and 68.

I also did a battery run down test of sorts. I started running the Sims3 on battery at 16:35, System went into hybernation at 18:31 with 5% left.

 

Which battery is it that you have - i.e. how many cells?

 

The 8 cell. However running the Sims3 is very stressful on the CPU, graphics and (I imagine) the hard drive. Last night I tried a different test but fell asleep before it could finish.

I'll be running 720p video full screen until it shuts down. That should last considerably longer. After that I'll try a less stressful surf-email-spreadsheet on wireless test.

 

krap101, my settings also go as low as .8750V on all 4 multipliers but the reading is . I don't get the same temperature as you though. My idle is about 34-35 degrees Celsius. Underload temperature goes up to 60 degrees Celsius...How'd you get yours so low?

 

Don't pay attention to the temps on the screenshots. Those were just to show the settings/multipliers/voltages. After surfing for about an hour, my temp right now is 51/55 for both cores. Idle is probably around 35-40, which is probably due to ambient temps.

 

Last night I did the surf, etc. test. It also included downloading and watching a 75 MB 22-minute video.

Result was 4 hours of life until shutdown with 7% left. The amount of time showing as left fluctuates with your activity. It took a big hit while watching the video, for example, then went back up when I resumed other activities.

Started: 19:19

3:39 left at 19:24, 97% battery life
3:36 19:34, 93%
3:14 19:42 90%
2:47 19:49 87%
3:42 19:57 84%
3:27 20:09 80%
3:23 20:19 77%
2:54 20:39 69%
2:57 20:48 66%
2:00 21:12 56%
2:22 21:13 55%
2:03 21:34 47%
1:38 21:50 41%
1:40 22:01 36%
0:54 22:16 30%
1:08 22:21 27%
1:06 22:29 24%
0:53 22:40 20%
0:50 22:45 18%
0:27 23:07 10%
0:21 23:14 7%

23:17 automatic shutdown

 

so with the p8600 is there no way to actually go below 1v? I haven't seen any temp decreases

 

If you look back just a few posts in this thread you will see that I did go below 1 volt, and temps dropped about 5 degrees Celsius

 

And I got to 0.95 V and dropped 10.

 

yeah, I read the posts, but I'm still having the same problems that other members mentioned. On the 14z, I can't get the voltages (according to CPUz) to go below 1v, no matter what RMClock is set to. I was just curious to see if anybody had found a way around this yet

 

I don't know what the problem is; if you look at my post you'll see I put the CPUz screenshots on and they show it below 1 volt. I did nothing that the undervolting guide did not tell me to do.

 

hey has anybody undervolted the 14z with the t9550.. if so please tell me what voltage you are using

 

It doesn't matter what voltage they are using, you might get completely different results.

Read about undervolting here

 

Just read the original thread - now I'm psyched. I probably won't do a full undervolt to start with, but I'll likely drop that .1 V right off the bat and test that.

 

I was actually able to get my voltages down to 1V for all my multipliers. I'm probably going to run stress tests again to make sure as I didn't get any BSOD's but I might have gotten an error and not seen it?

Is there anyway to undervolt below 1V? Just like posts #43, #44, #47, and #50 there is no voltage lower than 1V in the drop down menu in RMClock. I'm running a T6500 processor if that factors in.

 

Make sure you check "Mobile" and "Apply these settings at startup" in the advanced cpu settings.

 

yeah. I'm gonna second the 1V barrier.

all the settings are the same but it just wont budge.

 

Well that's the perplexing part. "Mobile" is selected and "Apply these settings at startup" is checked. Yet I am still unable to undervolt any of the multipliers lower than 1V. I am running on Visa 64bit on a T6500 processor if that matters.

 

I didn't seem to have any problems going below 1v on my 14z/P8600. However, I do notice from the CPU info screens previously posted that DFFS is disabled. If I turn DFFS off, then my voltage also locks to 1v like the others. With DFFS on, I can get 0.875V at 6x, reported both from CPU info and CPU-z. The undervolting guide said to turn it on, so I never thought to turn it off.

Now, here's where I'm having problems. DFFS is required to enable Super-LFM, so right now my Super-LFM is unchecked in the profile. If I attempt to turn it on, regardless of the FID/VID settings, the system will hang as soon as it enters S-LFM. Has anyone been able to successfully get S-LFM to work on a 14z/P8600?

 

Hmm.. that's strange. I have the exact same processor as you and I'm able to set all of the multipliers to 0.9250.

Did you download and re-patch the Vista 64-bit file in the RMclock folder? If you didn't, that might be the problem.. You can find it at the bottom of the undervolting guide.

 

THANK YOU. That finally fixed my problem.

 

Wow. By "hang", you mean that the computer just freezes right...? That's what I'm getting...

 

How much does the SLFM actually lower the voltage anyway? (for the record I have the same error, but I don't really care because I'm now at .925V for my highest multiplier, and .875 for my lowest. P8600)

 

There's no lower voltage for it to go, but it can halve the FSB. So for those idling cases, it can reduce the heat and power drain. For normal use, it likely won't help at all, so it's not really important. Just odd that the system will freeze and wondering if anyone had any success.

 

After reading this I was able to get my voltage down to 0.95, at .925 I get an error message every time.

The weird thing is that from .975v through .925v the temps were 60/62, so I guess the only savings there would be the minimal battery time I would gain. The rest of my voltages/temps looked like this; 1.000v 67/69, 1.025v 68/70 and stock 1.125v 73/77.

I'm just going to leave it at .950v, pretty easy way to save 13/15 degrees

 

Same. My minimum for the highest multiplier is .95V, now I'm testing the other multipliers.

 

So if I understood you correctly, I should be able to under volt past 1V if I turn DFFS on. Here is where I run into yet another problem...

I am unable to check the box that enables DFFS because it is grayed out. Also I do have the right CPU type selected so I don't think that should be the problem. (I was actually thinking the free version of RMClock doesn't have this option enabled while the paid version does.) And if it matters, my RMClock doesn't show anything for SuperLFM and IDA.

See the posted screen shots to see what I am presented with.

 

ah!!...atlast someone with the same problem as me...i tried everythin..i cudnt get it below 1V...i even tried undervolting using Crystal CPUID...didnt work either...
there has to be someone who knows how to solve this...

 

The issue I was addressing was for the 14z/P8600, where you can select under 1V in the profile, but it is never applied (limited to 1V) until you turn on DFFS. I don't have any other processor, so I can't speak for their behavior.

Looking at the other posts, it seems like the T6500 does not have these processor features. There was someone with a T6500 that managed to get under 1V on post #42. A close look at their CPU info screen shows TM1 and TM2 disabled. You might try turning those off.

 

Undervolted my Studio 14z, P8600. Except for the highest multiplier (9.0x), I could set all other multipliers to the lowest possible VID of 0.8750V. The highest one was throwing errors in ORTHOS, so I had to set it to 0.9250V.

 

Wow, I leave for a day and you guys accomplish so much. I think this solves my problem. I'm checking now.

 

OMFG YOU GUYS ROCK. (Especially you ikoiko ) FLIPPIN AWESOME

 

both of those are disabled for me, as well as the DFFS box being greyed out. @nilosays, I have the same t6500, but no luck with the 1 V barrier. I've done the 64 bit patch and everything.

 

Try resetting rightmark with that registry thing in the rightmark folder and see if it helps.

Also have you looked here?
http://forum.notebookreview.com/showthread.php?t=235824

 

Stable for 30 minutes at .925

stable for 20 at .9125

bsod at .9, so yeah, .925 seems like a good place to stop... doing overnight stress test

Running 10 minute test at 1.0375 to find out the temp diff.

Temps for the cores at 1.0375 are 69C and 74C (the room is kind of hot, I'd say close to 80F)

Temps at .925v after 15 minutes are 64C and 69C

Time to start on lower multipliers.

8x multiplier is at .8875 at 55C and 60C Stable for 90 minutes.

 

Just wanted to share my undervolting experiences with everybody.
My 14z has a 2.2 GHz T6600 processor.
The before undervolting the default VID for the highest multiplier (11x) is 1.15 V and the lowest multiplier (6x) is 0.925 V.

Being a risk taker I started undervolting by setting all the VID to the minimum.

After using RMClock to undervolt, I have all my VID set to 0.925 V, the lowest selectable VID for all the multipliers. That is an impressive 20% voltage drop while maintaining full speed performance!
I tested the system stability at max CPU speed for 7 hours under full load in Orthos.

Because the CPU fan speed is determined by the CPU temperature there isn't a good way to determine the absolute core temperature drop for a fixed fan speed, but after undervolting under full load the CPU temperature (as measured by CoreTemp) has dropped from 72C to 59C. I notice that under full load after undervolting the CPU fan doesn't have to go full speed all the time and when it does the CPU temperature drops to 57C. The ambient room temperature is about 23C.

Based on the textbook formula for dynamic CMOS power consumption
P = C*Vdd^2*f (C = capacitance, Vdd = supply voltage, f = operating frequency) and the Intel specified maximum TDP of 35 W, undervolting from 1.15 V to 0.925 V at maximum CPU frequency will drop the power consumption by 35% to a maximum TDP to 22.6 W (below the 25W TDP of a standard P8600).

Here some screen captures.
CPU-Z:

CoreTemp:

Orthos:

RMClock:

 

I also get 0.875V at 6x and 0.925V at 9x, so those look like common numbers for a P8600. I just did the auto levels for the steps in between, but now I think I'll go back and fiddle with the 7x/8x since you guys are getting lower values than the auto ones.

 

yeah, that's what I've gotten also with my p8600.

 

I'm in the exact same situation as karan1003 (post#91) - unable to undervolt any lower than 1V on my T6500 processor. This is starting to make me believe that the T6500 processor is locked at 1V cause all the other processors don't experience the 1V barrier or are able to get past it. However, there is one person (nilosays if I'm not mistaken) that was able to undervolt below 1V on a T6500.

Maybe I should try using a different software instead of RMClock? Maybe CrystalCPUID or CPUGenie? By the way, when using CrystalCPUID there's an (X) next to some of the voltages. Does that mean I am unable to use those voltages?

 

My P8600 cannot go under 1v either.

 

did you check the DFFS box? (advanced cpu settings, bottom box)

 

That box is greyed out for my T6500