NotebookForums.com › Forums › General Notebook Discussions › Notebook Guides › Simple Undervolting Tutorial - Pentium M Dothan
New Posts  All Forums:Forum Nav:

Simple Undervolting Tutorial - Pentium M Dothan - Page 47

post #921 of 1202
Thanks Mr. K6,

So you don't adjust voltages at all on the intermediate multipliers? It just seems odd that they are all left so high, but I suppose if the processor is almost never running at those intermediate speeds, it shouldn't make a difference, right? Do I understand that correctly?

And if so, why do you suppose intel bothered to make so many intermediate steps in the chip? Any ideas?

FWIW, I am just starting to get my feet wet, thanks for the help.

Vespadaddy
post #922 of 1202
No problem, you have to learn somehow , lemme know if I dont explain something clearly.

Quote:
Originally Posted by vespadaddy
So you don't adjust voltages at all on the intermediate multipliers? It just seems odd that they are all left so high, but I suppose if the processor is almost never running at those intermediate speeds, it shouldn't make a difference, right? Do I understand that correctly?
Exactly

Off hand, I dont really know why intel would bother. Maybe there are instances where it would function, but I just cant think of anything end user that would say, use only 50% CPU constantly, and if that's the case, why wouldnt you want to devote full speed to that process, you know? Overall, it might have been easier when incorporating speed step technology, specifically in scaling the CPUs. Reason being, if you have a 2.0GHz that runs at 15x133 and a 2.13 that runs at 16x133, if you specify all intermediate multipliers in the chipset/BIOS, then you dont have to write a new code for each CPU, but rather just if it would be an intermediate or full speed multiplier (i.e., 15x is full speed on a 2.0, but intermediate on a 2.13). Sorry if that goes way to in depth, it could be far off from the truth, but it seems like a (partially) logical explanation. Someone with more info should chime in here
post #923 of 1202
Here is another odd question. I have a space heater plugged into the same circuit as my laptop. Is it possible that the elctrical surges from the space heater cause voltage fluctuations that in turn cause errors when running prime 95?

This space heater does dim the lights when it cycles on.

I am running a Z70VA. 2ghz Pm, 1 gig ram. I have it undervolted to 1.116. I tried to run it at 1.084, and it ran prime 95 all night, but error'ed after 9 hours (which is also when I turned on the space heater).

Does this sound plausible, or should the power brick and internal circuitry smooth out any power fluctuations? What about the battery? Does the design of the Z70Va allow for the battery to serve as a sort of voltage sink, thereby offering smooth voltage, regardless of fluctuations in the AC current? I keep the battery in my laptop almost all of the time. It was in when Prime95 errored.

Also, does the warmer-running nature of an ASUS Z70VA tend to be less well suited for undervolting, due to the heat it produces?

Thanks!

Vespadaddy
post #924 of 1202
Thanx needledik.
My results are as follows:
Minimal Speed (800Mhz 6x):
Maximum CPU temperature @ 0.700V - 113-122 F
Maximum CPU temperature @ 0.988V - 122 F
Maximal Speed (1600Mhz 12x):
Maximum CPU temperature @ 0.956V - 127 F
Maximum CPU temperature @ 1.356V - 160 F
While reading the other comments i found my temps too high I have Toshiba M40-101 notebook with 1.6Ghz Dothan and ATI x300 video.
post #925 of 1202
Quote:
Originally Posted by Mr. K6
eh, NHC I havent really used. I use the previous version, CHC, which I find better and less of a resource hog.
I saw your post about NHC using an awful lot of memory (170 mb). Well, I just checked mine and it only used 3.4 mb memory. I don't know what was going on with your NHC...
post #926 of 1202
vespa: I'd recommend getting a surge protector anyway, you never know what could come down. Power surges shouldnt affect your CPU (the battery should kick in if power from the adapter drops), but try stress testing it in a different outlet/circuit anyway. My 2.0 instantly fails prime at 1.052v, but I run it at 1.116v everyday just for that extra buffer.

ruselok: Different manufacturers have different cooling solutions on their CPUs. Dell has a pretty good solution, so temps are low. Other notebooks that are smaller and have less of a cooling solution, so they will obviously run higher. However, Pentium M's are good up to 100C before they thermally shutdown, so even your stock 70C was within specs. What you should look for, though, is the drastic drop in temps that you got by undervolting (btw, report temps in Centigrade, it's what the tech world uses). You went from 71C to 53C which is an astounding drop, I think undervolting went very well for you
post #927 of 1202
Quote:
Originally Posted by vespadaddy
Thanks Mr. K6,

So you don't adjust voltages at all on the intermediate multipliers? It just seems odd that they are all left so high, but I suppose if the processor is almost never running at those intermediate speeds, it shouldn't make a difference, right? Do I understand that correctly?

And if so, why do you suppose intel bothered to make so many intermediate steps in the chip? Any ideas?

FWIW, I am just starting to get my feet wet, thanks for the help.

Vespadaddy
the intermediate voltages are interperlated (from the min and max) in a linear fashion unless Dmitri Besedin has changed RMClock but i doubt it... the voltages for the intermediate steps can be tweaked but the gain would be so marginal as to be virtually meaningless...
post #928 of 1202
Quote:
Originally Posted by vespadaddy
Here is another odd question. I have a space heater plugged into the same circuit as my laptop. Is it possible that the elctrical surges from the space heater cause voltage fluctuations that in turn cause errors when running prime 95?

This space heater does dim the lights when it cycles on.

I am running a Z70VA. 2ghz Pm, 1 gig ram. I have it undervolted to 1.116. I tried to run it at 1.084, and it ran prime 95 all night, but error'ed after 9 hours (which is also when I turned on the space heater).

Does this sound plausible, or should the power brick and internal circuitry smooth out any power fluctuations? What about the battery? Does the design of the Z70Va allow for the battery to serve as a sort of voltage sink, thereby offering smooth voltage, regardless of fluctuations in the AC current? I keep the battery in my laptop almost all of the time. It was in when Prime95 errored.

Also, does the warmer-running nature of an ASUS Z70VA tend to be less well suited for undervolting, due to the heat it produces?

Thanks!

Vespadaddy
yeah i've always wanted to know how much the adapter, etc. act as a conditioner... i've owned nothing but notebooks since my first pc over 10 years ago and i've never had any power problems of any type and i keep them on from the time i get up til i go to bed so i'm assuming they do a pretty good job of conditioning...

btw, i've been into very high end audio for many years (well actually WAS, kinda gotten away from it for some time) but i've been told by several of the maufacturers that generally speaking, the more expensive the product, the better/more stable the built in power conditioning... high end audio equipment has an unreal amount of conditioning - not even a ripple gets through, one of the keys to great sound...

several years ago i had a lightening strike (not uncommon in Atlanta lol) and it took out my phone, ans machine, etc. but didn't hurt any of my more expensive electronics...

space heaters draw a lotta amps so the circuit could be overloaded but i'd think the circuit breaker would flip first... but since we're talkin' about such small voltages on the pc, it might be the culprit but i kinda doubt it... just a guess on my part...
post #929 of 1202
Mr. K6
Yes, 18 C degrees drop for maximum performance is a very good result. But...when i just bought my notebook this year the temperatures were much lower (about 10 C degrees lower in idle and top load). Last time i started doom3 the temperature was around 80 degrees (air temperature 25 C degrees) I don't know what's wrong with my notebook. In service centre they just told me that the temperatures are within specs =( In Centigrade my results are as follows:
Minimal Speed (800Mhz 6x):
Maximum CPU temperature @ 0.700V - 45-50 C (cooler is running periodically)
Maximum CPU temperature @ 0.988V - 50 C (cooler is always running)
Maximal Speed (1600Mhz 12x):
Maximum CPU temperature @ 0.956V - 53 C
Maximum CPU temperature @ 1.356V - 71 C
post #930 of 1202
I used CHC (NHC harddrive control is buggy right now - Beta) to undervolt my machine. Bios A05 without fan control.

6x @ .70 got 30-32C (surfing) from 30-33C
15x @ 1.068 got 46C (Prim 95 medium for 5.45 hrs) from 56C (playing Moh maxed out).

Sweet tweak. Should I up the volts a step from the current minimums for safety sake? Is 5.45 hours ok with Prim95 or should I go longer?
post #931 of 1202
Quote:
Originally Posted by tcat92
6x @ .70 got 30C (surfing) from 30-33C
My gosh... how can you have 30C with your config?

I can never go below 46 C at 6x @ 0.716v...

Is there something wrong with my notebook?
post #932 of 1202
That's what CHC is telling me my cpu is: 31 C. When window resumes from standby, the cpu is 28C. Fyi, my ambient temp is 79F.

Any other XPSers out there with similar temps?

I was getting 31 or so even before undervoltat 6x so it didn't make any diff. Just the 15x. 10C drop!
post #933 of 1202
How could be such a huge difference between the same processors???

Could somebody explain this?

Also the XPS2 has that huge GPU inside which probably heats up the whole system even more... these temperature results are really weird.
post #934 of 1202
Quote:
Originally Posted by bveld
My gosh... how can you have 30C with your config?

I can never go below 46 C at 6x @ 0.716v...

Is there something wrong with my notebook?
I wouldn't worry about your 8104. Mine idles\operates at 46-48C and games at +-72C and I've seen other 8104s in this "ballpark". I use RMClock (and have used CHC to the same settings and near-same results) at 6x 0.700, 15x 1.100. Yes, his XPS Gen temps are extreeeemly low for a Pentium 760. You should check for other XPS Gen temps to see if they're all in that range. His CPU temps are more akin to my own HDD temps.
post #935 of 1202
Quote:
Originally Posted by MrMsyvc
I wouldn't worry about your 8104. Mine idles\operates at 46-48C and games at +-72C and I've seen other 8104s in this "ballpark". I use RMClock (and have used CHC to the same settings and near-same results) at 6x 0.700, 15x 1.100.
Your undervolting is very nice indeed. Actually I didn't try 0.700v because I wanted to leave some "safety margin" (as I said I have 0.716 now).

I haven't undervolted at 15x yet, maybe next week. 1.100v is very good. I was going to aim at 1.148v as it was stable for most of the 8104 owners.
post #936 of 1202
I just finished, now Im running-
min-6x @ .732v
max-15x @ 1.164
Seems to be running smoothly, I will post any drops in temperature later...
Nice guide
post #937 of 1202
bveld,

boy you had me scared for a minute. I thought my sensors were wacked. It seems my temp are normal based on other posts and I9kfangui.


http://www.notebookforums.com/showth...highlight=temp
post #938 of 1202
Quote:
Originally Posted by tcat92
bveld,

boy you had me scared for a minute. I thought my sensors were wacked. It seems my temp are normal based on other posts and I9kfangui.


http://www.notebookforums.com/showth...highlight=temp
I guess the XPS is just very good with cooling. But I still don't understand why mine is running so hot.
post #939 of 1202
Quote:
Originally Posted by bveld
My gosh... how can you have 30C with your config?

I can never go below 46 C at 6x @ 0.716v...

Is there something wrong with my notebook?
There are many things affecting average temperature readouts.

Ambient temperature is very important. I have my laptop running 4 degrees cooler on average at home (due to better room air conditioning) than at work.

Just recently, by applying AS5 to the heat removal assembly I managed to increase temprature readings by 3-5 degrees. The reason? ATI chipset was not properly attached to heatpipes and hence did not contribute to temperature sensor all the way (was not cooled enough too).

It also depends on location of temperature sensor and whether it is working as intended (i.e. not defective). Defective sensor may report lower (or higher) than actual temperature.

Different type of memory (and number of sticks) contribute to temperature in laptop case differently too. Other factors are: how high is your laptop raised over table top, is your wireless on, etc.

Research average temp readings from many users of the same laptop model and you'll get a ballpark number that's relatively safe. Anything significantly lower or higher is a problem (low temperature reading may be a sign of trouble too, see my example above).
post #940 of 1202
9300/XPS2 have very good cooling solutions. My laptop is idling at 27C as I write this (room is 25C). If temps are rising and you've had your laptop for awhile, it might be worth it to clean out your fans and heatsinks, dust might have built up^^
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Notebook Guides
NotebookForums.com › Forums › General Notebook Discussions › Notebook Guides › Simple Undervolting Tutorial - Pentium M Dothan