Replacing TIM

[Admiral Huddy]

When I built my currrent PC 18 months ago I didn't have any AS5 hanging around so I just used the TIM pad that came with the Artic Freezer.

I couldn't be bothered to change it but since it was a wash out this weekend, I thought I'd replace it with AS5.

The results were quite surprising... the CPU temp was down 5oc to 40oc on full load.. that's on an overclocked X2 4400+ with both cores running an instance of F@H..

:)

[Stan_Lite]

The stuff on the AFs is usually pretty good - I've only noticed a drop of 1-2C when I've replaced it with AS5 (maybe I'm not doing it right :p ).

That's a nice temperature for an X2 4400 at full F@H load.
I'm assuming you're using something like MBM or PC Probe or Speedfan for reading your temperature. A lot of people insist these applications are wrong and you should use the "Incredible and infallible" Core Temp as it "reads the true temperature of the cores". I personally don't trust Core Temp (any application which can't read the Vcore properly is not to be trusted to be reading temperature correctly, in my opinion). Many people have argued with me that it is more accurate than these other applications but when challenged to prove it, the best they can manage is to quote the developers claims.

Just because it reads the temperatures as being higher than other applications and is claimed by the developers to be more accurate, doesn't mean it is correct. If anyone can successfully prove that it is actually reading the correct temperatures, I will happily eat my words - until then, I will continue to use other applications (which actually manage to read the Vcore correctly).

Excuse the off-topic rant :embarassed:

Stan :)

[Daz]

Surely it's reading temps from the same api(s) anyway?

[Admiral Huddy]

Surely it's reading temps from the same api(s) anyway?


I would have thought so too.. APIs reading from the same settings as shown in the BIOS?

Yes used Asus probe and Everest.

I have two Scythe 120mm fans coming in from the front of the case and one at the rear and one on the roof. I did this about a month ago... I've replaced the fans that came with the case and run much quieter and cooler. I wasn't expecting such a big drop in the CPU when changing the TIM. Although, looking at the surface, it was clear that it needed replacing.

[Garp]

Surely it's reading temps from the same api(s) anyway?

Indeed.. it'll be the same info. The only time a note of caution is needed is in pointless comparisons with other PCs. Unless its the same motherboard any comparison is null because no two motherboards ever seem to be calibrated the same way. The main thing I do with every rig is figure out what my typical idle temp is and compare the full load temp. My shuttle is doing reasonably well in that it seems to show only a 8 degree increase in temps between idle and full load, and thats without the CPU fans going up to crazy speeds and noise level. My GFX card makes the most noise in the box, and even thats nice under idle. On my other identical shuttle that just uses its integrated gfx card I have to double check to see if its even switched on :D

[Davey_Pitch]

I once set up my pc and forgot to put any thermal paste on it, and was confused when my pc was shutting down after a few minutes. It took me 15 minutes to realise what I'd done, and a quick application later everything was lovely and smooth. It's a mistake I'll never make again! :)

[Admiral Huddy]

I once set up my pc and forgot to put any thermal paste on it, and was confused when my pc was shutting down after a few minutes. It took me 15 minutes to realise what I'd done, and a quick application later everything was lovely and smooth. It's a mistake I'll never make again! :)


You were lucky you didn't bork it lad! :D

[Feek]

I thought this thread was going to be about the Speaking Clock.

[Admiral Huddy]

I thought this thread was going to be about the Speaking Clock.

I did think TIM was still going... didn't he have a stroke?

[Feek]

I hear it was his third.

[Stan_Lite]

Indeed.. it'll be the same info. The only time a note of caution is needed is in pointless comparisons with other PCs. Unless its the same motherboard any comparison is null because no two motherboards ever seem to be calibrated the same way. The main thing I do with every rig is figure out what my typical idle temp is and compare the full load temp. My shuttle is doing reasonably well in that it seems to show only a 8 degree increase in temps between idle and full load, and thats without the CPU fans going up to crazy speeds and noise level. My GFX card makes the most noise in the box, and even thats nice under idle. On my other identical shuttle that just uses its integrated gfx card I have to double check to see if its even switched on :D

PCProbe shows my idle temp to be 24C and full load temp to be 51C - which seems like a big jump, unless you take into account that it is an overclocked, quad core radiator. Coupled with the overclocked, 8800GTX radiator below it, it's hardly surprising that the inside of the case gets extremely toasty. As a consequence of this, I prefer to leave the side off the case (especially at this time of year) - the Akasa Eclipse may be a big case with loads of space inside but the airflow must be truly horrendous (it's very neat inside, before anyone asks - I made a special effort with it). I really must get a case with better airflow - any suggestions anybody?
It must be big enough to take an 8800GTX (it's surprising how many cases aren't).

Stan :)

Edit: :rolleyes: @ Feek for that truly terrible joke.

[Admiral Huddy]

^ I love my Kandalf :)

http://www.huddysworld.co.uk/PCHardware_BlogPage.shtml#Upgrade_-_A_case_study..