So i got my new ES laptop with ES 3920xm cpu and twin 7970m.
Did a ton of bitcoin mining with it, since AMD gpus are good at that.
CPUID Hardware Monitor shows only the thermals for the first GPU. I thought some sensors were missing. But then i instaled the HWINFO64, and here i have all the sensors.
Now i knew before i got it that its design is so that the second GPU shares the fan with the CPU. I sayd its not so bad since the second cpu will not be used fully as the first.
Then i started the miner and watched the temperatures.
![]()
Is this even remotely normal ?
Here is the disposition i have with the laptop:
![]()
![]()
Will test it in gaming to see what i get.
-
Diablo 3 Crossfire on, no fps limit, 65/85 temp for the 1st and second gpu, utiization between 43 and 55% for both gpus.
Diablo 3 Crossfire Off, first gpu game, second gpu bitcoin mining temp 73/103, utilization 98 for the first and also 98 for the second allthought it shows after some minutes utilization 0 for the second, probably due to the mining technique -
curious what performance do you get out of your 3920xm ES? When you run prime what speed does it drop to with all 4 cores maxed out after 5 mins?
-
103 is way too high. You should try repasting that GPU.
-
curious what performance do you get out of your 3920xm ES? When you run prime what speed does it drop to with all 4 cores maxed out after 5 mins?
What exactly do you mean with this ? -
when your CPU is under full load for over 5 minutes, what Ghz does your turbo boost stay at? Just curious on how efficient engineering sample of 3920xm would be, having owned a few ES chips in the past.
-
3591 mhz all 4 cores, temperatures reach all the way up to 100 cesius 86/101/103/101 temperatures max per core
Wait, i need to let it 5 minutes, the freq seems to drop a bit
The freq is not stable, going through 3392 3492 3592 mhz
With cpu load, the second gpu fan also kicks in, it reduced the gpu2 temp to 35 celsius in no load -
wow thats pretty hot, I would repaste your CPU and GPU. p150em temps only reach 88c after 5 minutes of prime @ 3.8ghz constant with 3720qm, so you should be getting even lower temp.
-
moral hazard Notebook Nobel Laureate
Is it really an ES laptop?
Or do you just have an ES CPU and GPUs but the notebook is OEM? -
The laptop is ES, cpu is ES and gpus i dont know about them, perhaps they are ES too.
-
Should be ES book and CPU but OEM GPU.
-
Try repasting. Also copper heatsinks should help a lot.
-
moral hazard Notebook Nobel Laureate
Wow I can't get over the fact you have an ES notebook, care to share the source that you got it from?
How was your hash rate while mining?
Mining seems to heat up my GPU a lot more than any other software.
As far as temps go, I could check to see how well your thermal paste has spread on that GPU. Also maybe you could check how fast your fans are spinning with hwinfo, could it be possible that you could use an EC update because your fans are not spinning fast enough?
Does your notebook have the FN + 1 key combo for max fan speed? -
Minig hash rate is fluctuating between 250 and 310, I don't know if that is normal. 75 to 80 degrees for the 1st gpu, 103 max for the second.
Fn+1 doest seem to work, ut fan speed shows as 100, but for the 1st fan, not always.
I kept it 18h a day mining with both gpus, and when im home and playing something I played only on the 1st gpu, cause 60fps was all I needed, and kept the second gpu to mining. So the second gpu mined pretty m0ch 24 7.
I th1nk ill returnit though. 103 degrees for the second gpu and for the cpu as well, its not normal. Repaste, didn't try it so far.
Sent from my GT-N7000 using Tapatalk 2 -
FYI - Downclock your memory - it has ZERO to do with your mh/s. This will drop your temps too. Your gpu will be running at 99% while mining, hence these temps.
With the savings in temps on your memory - you may actually be able to overclock the core which is where you will actually see increases in mh/s
Edit - just saying to do the above as well, not saying repasting isn't also part of it. But those stock memory clocks aren't helping you mine, they are only helping create heat! -
Does the P370EM have the same issue with the 7970m's like the older P170EM and the P150EM with enduro? Im actually in the process of making a decision on 4 different laptops (yeah laugh there not all going to be for me, i wish
The unit i want is the P370EM and i'm just trying to figure what config i want, i want to get the XM processor but if i get it with 680m SLI it would put me over my budget for that laptop so my other option is the 7970m's, Then i'm ordering a P170EM and (2) P150EM's
-
moral hazard Notebook Nobel Laureate
I get 370mh/s with a single 7970 @ 1030/1200.
Temps about 80C with the back raised and it's summer now.
So there is a small problem somewhere in your notebook, IMO. -
I have a 7970m CF ; Temps on the second GPU , can get quite high for me - 95c highest i have had.
What cooler are u using?
These things are so beasts, so much power in so little space - but still 100 c is really hot.
Not sure how effective the vents are on these - is could be possible that the cooler is actually hindering the process of heat disspertion forcing hot air in, but thats unlikely, but maybe worth trying without the cooler.
These laptops are really at the cutting edge of tech, and to get 11,000 3dmark11 on a portable 17.3 laptop would always have some drawbacks.
Anyway, let me know how you get on , and if re pasting helpsWant to try and lower mine too
pj -
-
Thank you guys for the quick responses
-
-
Anyway to edit the vbios to lower the clocks?
-
moral hazard Notebook Nobel Laureate
Svl7 seems to be able to mod the vbios (clocks and voltage).
-
Sent from my GT-N7000 using Tapatalk 2 -
So what was your solution, TC? Are you still using your laptop with over 100C GPU temps with no problem?
Received mine today, that's just INSANE!!! The temp for the 2nd GPU climbs so high, just afraid to try any game out there
Did a repaste, no help((( any help, fellas?
Used the FN+1 and oced the CPU to 4.1Ghz using Intel's XTU. Thought it was the same CPU/GPU HS problem, but even at stock the gpu hits 105 and stays there. 12.11 beta 8 drivers -
-
I bet the sink is not making full contact with the gpu die. You guys should try a retention mod. Just as a comparison, in the M18x with overclocked 7970m xfire the max temps were in the 77 C range. It seems Clevo is doing a poor job with fitting it's sinks lately.
Sent from my GT-N7000 -
the temps seem fine now, all i had to do was to push really hard on the GPU2 HS around die during benchmark so the paste really sticks to the hs. turn the unit off for 5-10 min and voila)
it's clearly the hs that makes poor contact with the GPU's die, that's all! -
I knew it! What you can do to fix this is add a 1-2mm plastic washer to increase the tension for each screw and the sink should make much better contact. Also make sure it fits properly on the card and that something isn't obstructing it. -
Does adding the washer actually fixes the high temps ?
mine are hitting 103c on games like ( Farcry3 - bf3 - sleeping dogs )
i just cant enjoy playing any game with these temp on the 9370.
can some1 do a step by step DIY ?
i really dnt have a clue when its comes to hardware and fixing. -
Support.3@XOTIC PC Company Representative
-
When I ordered it iI requested ic diamond compound ,i even went
further and asked my reseller to keep the laptop few more days to make sure
Everything is working gr8 , thats how am paid back i guess .
some have already did a repartee with same bad results so Not sure what to do now , -
Moudine
Hey, got the same problem with 7970M CF on P370EM and i found a solution that worked right out! I detached the GPU HS from the overheating card and the cooper part that covered the GPU die was not flat enough (you could just put a ruler to the surface to see if it's curved) In this case neither paste/cooper hs helps. I just bended it to a proper level with my hands (nothing difficult) applied a TIM and voila 85C instead of 103C + throttling)
Hope that works for those who have this issues -
Hey redline , thanks for the info.
I opened the back cover and started far cry 3
I monitored the temp and as soon as it hit 95C +
I pressed the heatsink using my hand with a pc of
Paper and it actually rapidly dropped the temp to
Around 85C ... But when i removed my hand temp went up
Again so ur idea does works.
I never tried to remove the heatsink , i also looked
At it and it appears to be straight .
Removing the heatsink on the die requires removing
Only the 4 screws there ?
If i removed it will i need to repaste it or the ones there
Can still work ?
Reason am asking is that i work far away ,
In africa and thermal compounds are not sold here. -
Moudine
wow, Africa) anyways you can't tell if it's flat or not without removing it. 1mm gap could cause the heat to jump. I guess it's the GPU on the right hand side, the one that share the HS with the CPU, correct? In this case you will have to remove CPU HS as well. If the one on the left side (primary GPU) then it's just 4 screws got to be removed. I would recommend buying a TIM somewhere in order to get a perfect layer but if there's no place to buy it try this:
Detach the hs, bend it to the proper way from the sides, attach it back, turn the laptop run any game for 1-2 min (let the paste spread by itself), turn it off for 15 min! And you're ready to go)
Hope that helps -
Did u mean i have to bend it from the sides only?
Left and right of the hs ? How much is enough
Sorry for asking a lot but in case i mess up am screwed lol -
yeah, just a little pressure from the sides is enough. if you have a ruler just make sure the surface that touches the GPU die is flat after you bend it.
-
Meaker@Sager Company Representative
Also make sure no thermal pads have folded over, this can also create a gap between the contact area and the core.
-
okay guys good news
... my temp has dropped around 10C .
i ran a game and kept pushing on the HT for like 1 min then turned off the machine,
i noticed there was a small gap on the HT around screw 4 so i used a small metal plate and added pressure
on that point which leveled it a bit...
i played far cry 3 for about an hour didnt get higher than 87 and bf3 max 82 so i thinks am in acceptable range of temp.
now just hoping results last and the HS wnt loose out again.
thanks all for the help really was gr8.
i still think of wat my reseller told me " clevo and sager assures that these temp are normal for this pc with this config "
103C sounds normal for them ! -
will be testing more games and sharing the results
-
Moudine, glad to hear that! Yeah, those temps look normal
-
On bf3 I just locked the frames @ 65fps with the .cfg file. Lower heat and a much more playable game.
-
Thanks redline . i ordered some thermal and copper pads ill try to seal up the gaps between the fan and radiator
perhaps i can even manage to lower the temp to mid 70's ( finger crossed ).ill keep u updated. -
Hi guys one question GPU1 is the master or GPU2? I have a Clevo P370EM crossfir 7970m
-
I've just finished a repaste on my 9370 and the temps went from 88 C @ 15 mins to 81 C @15 mins on Kombustor. Same exact room temp, location, etc maybe 30 mins apart.. is that a normal gain? I noticed the previous paste job was not the greatest, plenty of air pockets and extremely dry/discolored spots. I just can't believe that paste could make such a big diff. On my desktop i gained maybe 1-3 C with arctic 5 a couple yrs back so this kind of took me by surprise.
Off topic: OP what's the cooler/stand u have the laptop on? -
All is not perfect laptop, but with a price point you can afford to improve the cooling
P370EM CrossFire GHz Edition
-
Meaker@Sager Company Representative
-
Thanks Meaker
i repaste the second GPU reseat the heatsink from 103C i am now down to 94C on Kombastor but is still to high any ideas what else to do? i repaste with prolimatech pk3. On kombastor GPU1 has VDDC 1.000v GPU2 VDDC 1.050V is it normal?
Clevo P370em ES with 7970m CF - High Temperatures !
Discussion in 'Sager and Clevo' started by Bytales, Oct 20, 2012.