are this temps look normal at all with DV9000?

hi guys, i was up last night and i got into reading about copper mod and from that thread i was pointed in the direction of undervolting thread.... well i decided to try it out

i read like the 1st 15 pages or so and decided to give it a try... downloaded orthos and HWmonitor, did a 100% cpu load for 20 minutes to find my max temp at load and i had 69*C highest on my core#1 (core#0 always seems to be cooler by 2-4*C almost all the time << is this normal?) i forgot to take picture of the 69*C cause i was soooo tired... well i got into RMClock, for some reason it did not reconize my T8300 but o well, it got all the multiplier and MHz right so i was happy. now as the guide says 1st drop about 0.100 VID off of the stock and run the test.... and so on.... well i fell sleep (did not wake up 45minutes after starting orthos to drop the vid down 1 notch...) so when i woke up this morning, 3 hrs later i did not have any blue screens or BSOD but when i looked at the temperatures.... they looked fishy. i know the guide says
*Advantages of Undervolting
- Cooler CPU (5 to 20c cooler)
- More battery life (15-30mins more)
- Less Fan noise & activity
- Longer overall notebook life
- Eco-Friendly



well i got 20*C cooler on the 1st run... i was hoping for it to be cooler by maybe 10*C at the end of all of the test runs... going from 69*C to 50*C max at 100% cpu load for 3 hrs is a huge drop don't you think? i want to believe i might just might have a cherry picked T8300... but i don't think i do. for some reason i don't feel lucky enough to get one. i am planning on spending my whole day at work trying out the undervolting thing and see what is my lowest VID that i can stay stable at and what temps i get. i am already at 1.0375V @ 50*C max for 3hrs of ORTHOS.

i am thinking maybe orthos is not stressing my cpu to 100% since when i put my hand down next to the fans by the cpu, it was blowing out cool at after being stressed supposedly at 100% for 3 hrs. (fan was running at full speed from the sound of it) you guys think ORTHOS is not doing its job, and maybe i should use Prime95 to put the stress on the CPU? or maybe HWmonitor is reading temps wrong.... i'm confused

here is the SS of the thing

 

If you really want some serious load give IntelBurnTest a go, it uses the same algorithm that Intel tests their chips at the factory with. However, i strongly advise you NOT to fall asleep while that is running. But you won't need to fall asleep - IntelBurnTest finds any instability in 20 minutes at most, that other programs can take up to 12 hours to detect. It also gets your CPU 10-15C hotter than anything else. Yes this is the claim of the program writer, but i have found it to be perfectly true.

Those are some really nice temperatures, and yes i think they are correct since you mentioned that the notebook fan blows out cool air. Congrats, you have one of the coolest laptop processors around.

 

This program(IntelBurnTest) is awesome.

 

Yes that max temp is normal for a T8300 and 2-4c difference between the cores is normal.

AMD's chips sometimes have like 10c difference between cores and can drop up to 30c in temps when undervolted

Yes but its not a worry. RMclock cannot detect newer revisions of your chip seeing as its an outdated software, it will just say unknown core.

T8300 is the best notebook CPU to undervolt. You should be able to run in on the lowest voltage possible.

Use OCCT or Intel Burn test. ORTHOS is practically Prime95

 

i downloaded Intel Burn test but for some reason i can't get the RMclock to work on this hdd no more. i took out the ssd and replaced it with my oem hdd and now i forgot what file i downloaded and put it into programfiles for RMclock... now i have to spend another 1/2 hr trying to figure out what was the file i nedded last night to get the RMclock running

 

Just install it again? Remember that you need to download the signed 64bit drivers if your running the OS in your signature

 

ahhh yes thats what i needed to download

man... this chip is really something i am currently at VID: 0.9625V and my max temps in HWMonitors for core 0/1 are 52*c/52*c and Core temp shows core 0/1 at 56*C/57*C this results are from running IntelBurn Test 5 times at standard stress level.

here are the results from alll of my tests so far..

Base max temps @ 1.1375V



2nd test @ VID: 1.0375V



3rd test @ VID: 1.0125V



4th test @ VID: 0.9875V



5th test @ VID: 0.9625V




now that i've got the ImageBurn stable on 0.9625V, i was thinking maybe i should do a stress level at Maximum and maybe run it 10 times to see if its pure stabel at 0.9625V and see whats a max temp on that. because i still don't think that the "Standard" stress level and running it 5 times really roves anything or makes it rock stabel. what do you guys think? if i can get this to run at Maximum with 10 runs, would that be a good stabel results? or i need to run it more than 10 times?

 

Set it to Maximum and let it repeat 30-40 times. Do it before you sleep or something.

If it passes then your good to go

If it doesnt, bump the voltage up by 2 steps and retest

 

I found that to be false.
I always test with ORTHOS for no more than 15mins, it doesn't take more than that.
I think some people just get carried away with it.


Also with ORTHOS you don't have to wait for it to finish running before you get results.

Again, with me running both programs (ORTHOS and intelburntest) for the same amount of time gave me the same max temp.

I think the programmer is just trying to advertise and make people convert from orthos.

While ORTHOS is old, it still is good for Core 2 CPUs.
maybe if you have a quad core then go for intelburntest.

 

ok, i will set this to do 30 times at maximum and see where it gets me. hum.. instead of doing at 0.9625V i'll take 1 more step down and try it at 0.9500V. i'm sure it probably won't passs 30 runs. if not i'll wake up in the morning and start the test at 0.9625V and see whats up.

 

welll i got blue screen within 1 min of running the test with 0.9500V lol... so i started the 0.9625V max stress and repeat 30 times at 2 am. now its 9:15am and it has only done 12/30 runs. man i could not wait for it to get to finish all 30 the test had been running for 7hrs straight and the temps never went over 44*C in HW Monitor and 48*C was the highest in Core Temp. so is this 12run max stress test for 7hrs good to say that this cpu is very very stabel at 0.9625V with only 44*C max temp? or do i need to run ORTHOS for like maybe 8hrs to compare the results with intel burn test and see if they are the same?

6th test @ 0.9625V for 7hrs/12 runs w/ max stress

 

Totally FAKE. Yes that's what i thought too - i built a computer for a classmate of mine last year. It's got a Q6600, a 4850, 4GB RAM, things like that. I managed to get the CPU up to 3.4GHz easily, stable in Orthos for 15 minutes, i left home. The 2nd day the system was bluescreening after 30 minutes in any game. I had to go back to him and lower the overclock to 3GHz, it's been running like clockwork since. Orthos can take up to 8 hours to find instability - i experienced it on my own computer - and IntelBurnTest reported it after 6 minutes. In the time i had been running Orthos i had programs crashing like crazy, but nothing was wrong as far as Orthos cared. I've had this happen several times - while running a stress test it doesn't fail itself, but causes other programs to crash. This is also a sign of instability.

If you really hate the guy that wrote the IntelBurnTest front-end, you can always use the Linpack command-line utility itself, which is written by Intel's engineers.

 

that Linpack64 looks like its in windows dos mode, and i can't work with that. for me its too complicated. that IntelBurn test so much simpler. all i do is choose the sterss level and hit start and i'm all done. its nice and simple and gets the cpu running at full throttle really fast

 

Sounds like the ram was not stable to me.
Did you try the "blend" setting in ORTHOS?

But the program doesn't show you the results untill it's finished running.
That is useless for me. With ORTHOS, you can just leave it running and keep increaseing the frequency untill you get an error.
Then when you get your final stable speed, go test with intelburntool.

I will try the command-line linpack.
But I doubt it would show me errors in real time like ORTHOS does...

 

In my classmate's system the RAM was unlinked from the FSB. And yes i was running the Blend test.

The latest version of IntelBurnTest also reports near-realtime results. The standard stress level takes only about 30 seconds on a modern CPU. If the results mismatch on the next pass testing is suspended and you are informed of this.