repasting and temps

I recently upgraded from a 2410m to a 2620m. (dual core). Most the programs I run are single threaded or maybe can use dual cores. I don't do any heavy encoding of any sort.

So after running prime95 for 15 min, i think the highest temp was 93c on one core. Is this unusual? I think the games i play use the CPU more moderatly. But if 93 is way to hot during a stress test I will be repasting in the way described here: (I spread the paste across the top of the CPU in my application)

Application

i'll probably research a little in the next week, but just wondered if anyone had any quick opinions. During prime95, I'd start off at ~3.4 and end up at 2.8-3.0Ghz within my 15 min tests. It never reached its minimum of 2.7ghz

basically pulling apart the whole computer takes a while and not sure its worth the effort.

 

Is this testing your cpu clocks? I wouldn't be surprised if it got that high but check it when you are playing games. Also be sure not to void your warranty by taking it apart, that never goes well especially with Dell...

IMO that is a little high and if you really want to repaste the gpu go for it but get a high quality thermal paste and if the temps still don't go down propping the laptop up so that the intake vent (wherever it is on your laptop) has more air to suck in AAANNNNDDDD If you want even more cooling get a nice 2/3 fan laptop cooling pad (powered by usb if possible) I think newegg sells the cooler master something u2 and thats a nice one for only 20 dollars

 

Well, prime95 is a really serious CPU bench program, i doubt you'll ever need that much of power from you CPU.. try to run a few 3dmarks/games and 85C should be your limit wich is quite attractive. while Tmax for desktop cpu's is 72.6°C and for mobile ones it's 100°C for my personal tastes i won't make it go more than 85C. Then repaste job isn't necessary. You're good to go)

 

thanks for your replies, I want to avoid throttling, I noticed higher framerates in a CPU dependant game I play, I think its ok. I did repaste the GPU as well since they all under the same heatsink.

The r2 of XPS 15 seems so cool I don't see how a cooling pad would help. Its mostly hot exhaust.

I'm using IC diamond 24, the manufacturer seems more into positive results than making a quick buck. I trust them 100%.

I'm mostly testing the temp, but as it raises I know the CPU clocks go down, so its kind of hand in hand.

I think I'll test temps under usual usage before I strip it down again. It wasn't that hard, just time consuming.

thanks again for your inputs.

 

I have a question for OP, i currently have a L502x with the 2410m and im looking to upgrade a few months down the road to a 2620m. Just wondering where you got the 2620m and if the switch was difficult? Sorry to hi-jack the thread a little.

 

no problem about hi-jacking. I bought a 'pull' off ebay, it said buy it now for $280. I offered $250 and he agreed. You can probably get engineering samples cheaper(? looks like not by much), but I didn't want to chance it. (most have success though). He was selling like 5. I'll search for a link..

Intel i7-2620M L2 4MB 2C4T SR03F 2.7GHz brand new - eBay (item 220799331683 end time Jul-14-11 09:13:54 PDT)

there's 2 left. I know he'll take $250 which is about what the engineering samples go for. Maybe he'll hold out for more on the last two, I don't know.

the switch required taking every single part of your laptop apart just about. Its why I'm not eager to repaste. Its not hard, Dell's repair manuals are excellent. It just takes time.

I've been looking for final production 2620m for a while now, took me a while to find one not engineering sample. If you could afford it I'd make an offer soon.

 

oh what the hell, just sitting here, I might as well repaste. I'll post results of high temp doing prime 95 for 15 min. 93c was high before.

If nothing else, it may tell people if the different application method makes a difference.

if it goes up higher, I'm going to be pissed

new temps after repaste: 79c highest. I'm pleased with the temp change, but it seemed to hover around its lowest state at 2.7GHZ using prime95 for 15 min. I'm pretty sure I was hitting higher clocks before. I guess ~10c is worth it though. I'm puzzled why it didn't stress the CPU as much.

From what I read there is time to let thermal paste set, I'm unsure if the IC Diamond 24 needs this though, its viscosity is pretty thick.

basically I'm confused now.

 

Good luck! As for me, my l502x arrived today so I'll be doing my own repaste job today or tomorrow. I'm planning to use the razor blade method with my AS5 (as per instructions on their website). We can compare our results on the temps thread, I guess. You might also find useful info in this thread with your problems, although I can't promise that for sure.

 

I'll post my results in that thread, thanks. I didn't mean to split the discussion