*OFFICIAL* M18xR1/R1/18 Owners "Are my CPU/GPU Temps OK Thread" - Ask Questions/Get Help Here

Excuse me , im new on this forum and im an proud owner of M18x
i want to ask, is it fine to turn the fan always at 4000~5000RPM? the heat just too much for me . aleardy repasted and in six months i have to clean up again

 

I just installed a 7970M in my R1 (thanks, Blaster). HWInfo64 only shows the single 'GPU Thermal Diode', whereas it showed several temperatures for my 6990Ms. Is that normal, and do I only need to pay attention to that single reading it provides?

 

That's correct, TurbodTalon. No worries... it's normal. There is no other GPU temperature reading that you can monitor.

 

Thank you, Mr. Fox. I have another question. How do you overclock it? MSI Afterburner shows no OC ability, and TRIXX shows the two sliders, but they're useless. I did a VBIOS dump with GPU-Z, but Radeon BIOS Editor (RBE) thinks it is corrupt. So my usual soft and hard OCing methods aren't working. How are you guys OCing the 7970Ms?

 

Have a look at the information near the end of post #1 and the content of post #2 in this thread: http://forum.notebookreview.com/ali...sink-7970m-crossfire-upgrade.html#post8568526. I think you will find some useful tips there, in addition to the answer to your overclocking question.

Obviously, you will not need any Crossfire tweaks with one card, but everything else applies to a single 7970M.

RBE and Trixx do not work... they are not compatible, unfortunately. There are a number of things that worked well with 6990M and 580M that not compatible with 7970M and 680M because they are new tech. You can use the vBIOS Patcher to mod your vBIOS to your own liking. With vBIOS Patcher and MSI Afterburner, you're not going to miss your old utilities.

 

Thanks again, Mr. Fox.

 

One last thing, don't flash the 7970M within Windows. Use ATIFlash with a bootable DOS USB stick only. ATIWinFlash may cause corruption or leave the vBIOS unreadable. It's recoverable by a system BIOS adjustment to boot with IGFX and discreet graphics active simultaneously in DOS mode, but you want to avoid that if for no reason than having to reprogram it is very inconvenient.

 

I had the 118 VBIOS, so I flashed to your 117. Now, do the 2D clocks and voltage need to match the 3D clocks and voltage? I ask because the card stays spun up all of the time, which causes the GPU fan to be running constantly.

That being said, what were the original 2D clocks and voltage?

 

The original 2D/3D clocks and voltage are shown in the GPU Shark screen shot in that thread I directed you to.

Are you on A04 or A05? I did not have any fan issues. With a single GPU you may not need to set 2D to 3D speeds and voltage. I had to or my Crossfire setup performed poorly if I did not.

 

I'm on A05, but I believe there may be other contributing factors to the fan being spun up. Is the single temperature reading on the core or memory? The fan is spun up just enough to hear, it isn't terrible. I'm using an Nvidia 75 watt heatsink (100 watt ATI is en route), and it doesn't put a ton of pressure on the memory chips. Firm contact with core though. I'll see what that new ATI heatsink does, and worry about it after that. The card never gets above 70°C, even when I'm playing StartCraft II on Extreme settings, but it idles at about 56°C. Thanks again for all the help.

Any ideas on a stable, daily-use OC for these buggers?

 

You're welcome, TurbodTalon. Mine were left stock most of the time because they did not OC with stability very well. 925/1400 never gave me any trouble, but it did not provide an advantage for daily use, just benchmark number chasing. The 7970M core and memory are already clocked high enough to do well with gaming and ordinary daily use. I dropped back to A04 because the fan activity is a little bit erratic with A05 regardless of what GPU is being run. The primary GPU fan runs more than necessary with A05. Once it spins up, it takes longer than appropriate for it to return to normal. A04 did not elicit that behavior. A05 was designed to fix the issue with 6990M overheating and thermal shutdown and the 7970M doesn't have that issue. I believe the temp sensor is the core. I think you probably will get better results with the correct heat sink. Lowering the idle temp could very well help with the idle time fan speed.

Hi there... and welcome to the community.

The maximum fan speed is approximately 3800 RPM. If you are attempting to force the fans at 4,000 to 5,000 RPM they don't turn that fast and will max out before reaching the speed limit you are telling them to run. That causes no harm to the fans, as they simply ignore instructions to run faster than what they are capable of running. You probably do not need to repaste again. If it has been 6 months, it probably just needs a good cleaning to get rid of the dust bunnies and lint in the fans and heat sink radiators. If you remove the lower cover and take the screws out of the fans, you can vacuum those out nicely. Just be sure to hold the fans still to they do not get damage from being over-revved.

 

impressive answer Mr.Fox ,
yeah as i suspected, it must be the dust that makes my alienware heats more,
for controlling the fans better manual or custom auto in hwinfo? any recommendation expect that?
i think my temps are now fine after i clean it. thank you so much :thumbsup:

 

Awesome news! And, thank you for posting that it helped you. +1 Rep

To be honest, unless you are running a wicked overclock or something, the stock fan profiles that are resident in the Compal EC table do a fine job. I use the stock fan profiles and, when I need to, use HWiNFO64 to force the fans to run at 100%.

 

If the gpus get friend do the laptop stop booting?. My laptop was working fine but after a while it got too hot and it did not want to turn on again. The specs are m18x r1 cx6990m and i7-2720QM

 

More than likely, one of the gpus are damaged. If you take them both out, the laptop should power on under igpu.

 

Your link goes to amazon.com

 

LOL, spam bot... "He" is banned now.

 

So, since I replaced my 6970s with 6990s and did clean driver install, my 3DV score is up 10% and I get the sameish increase with 3DM11.

However, I only get those scores if I raise the back of the laptop (2-3 inches) AND set the fan speeds to run at 3198 via HWiNF064. Otherwise the scores are halved as I think one of the card overheats disabling XFire.

I therefore have a problem with my fans?

When I click "Custom Auto" on HWiNFO and choose "sensor" it will only allow me control the fan on the first GPU, not the second. Why is that?

Also, I checked temps immediately after an hour of BF3, 1 card was reading 90 degrees and the other 40 despite the fact that the Crossfire logo was on during the entire game.

 

Hi guys. I have a M18X R1 with a crossfire of 6990m.

My temps are getting a little high (secondary GPU is at 100°C on very demanding games) so I think I'm going to repaste the Gpus et maybe the cpu.

Can you give me some advice ? Thermal grease or thermal pads ? What brand ? I've been out for a while now, I don't know what's good these days.

Thanks.

 

The best paste will have to be ICD 7 or 24 depending on what size of tube you need. I'll just leave the stock thermal pads in place.

Have you by chance attempted just to give it a good cleaning by removing the back cover and making sure no dust is on the fan or heat sinks?

 

No dust in the fans, I clean them every once in a while.

No, I had the cold boot bug a while back and I didn't get there was a problem right away, not until GPU 1 started glitching pretty bad.
So a DELL technician came, almost 6 months ago to change GPU 1 and we saw there was almost no thermal pads / paste left : it had completly melted away (eh, those temps, with no fans, no surprise).

So GPU 1 was changed but not GPU 2. Since then I've seen it getting hotter and hotter. At idle, GPU 2 is 10-15°C hotter than GPU 1, in charge, it's at least 25-30°C hotter ! While it's used less mind you (it's the secondary GPU after all).

I'm pretty sure it's the thermal pads / paste at fault here. Should have changed or repasted GPU 2 last time the DELL guy came but Alienware support didn't want to...

 

Hi guys,

I've had this system for about a week now, and it's been an absolute pleasure to work with. The only hangup I've developed with it are the CPU (2960xm @ stock) temperatures at both idle & load. Here in The Netherlands, the temperature is hovering around 5c-12c, with the inside of our home heated to a mild 17c. The notebook is on a wooden desk, propped up at the back about 2 inches by a book. I'm seeing idle temps ranging from 45c-58c and gaming temps around 85c. This seems a bit high to me. I'm not sure what heatsink the laptop has (it shipped with the XM chip) but I'm guessing it's the dual pipe? If anyone has any feedback/suggestions it would be greatly appreciated, just want to make sure she's running tip-top gents!

thanks,

jason

 

Hi Jason. There is nothing wrong with the idle temps. My system is idling at 63°C right now. It's fine.

Your CPU load temp of 85°C is not necessarily a big problem either. Is that a peak temp, or does it go to 85°C and just hover there the entire time you are gaming? What CPU clock settings are you using?

The first thing I would do is pop out the battery, remove the 4 screws for the bottom cover, slide it off and either vacuum or blow out the fans and heat sink radiators. Depending on the environment, this needs to be done at least once every 30 to 90 days. To do a really effective job of it, take the screws out of the fans and lift them off to the side to get the heat sink radiators clean from all the lint and dust. You don't need to unplug them from the motherboard, as the wires have enough slack to move them out of the way for cleaning. Hold the fan blades when you clean them so they don't spin at high RPM. That can damage the fans. Dirt and dust bunnies are the leading cause of heat issues with laptops and desktop computers. It is also amazing how many performance problems a simple cleaning can resolve.

After doing this, report back if your temps are lower. If they are still higher than you would prefer them to be, either a repaste of the CPU heat sink or changing your CPU settings in the BIOS would be a next step in lowering the temps. Don't mess with repasting the GPU heat sinks if they are not running hotter than normal. Best to leave them be if their temps are fine.

 

Ah, good to know those ranges aren't abnormal then =)

Load temps range from 55c-85c while gaming. While doing video encoding/editing @ 1080p in Sony Vegas, the chip sees 70c-85c, with a median temp of 80c or so that it hangs around.

I'll do a strip down and clean of the fans & heatsinks tomorrow and get back with new readings.

Another issue I've noticed, though this may be the inappropriate thread...the 2960's turbo clock frequency. It seems hell-bent on not going over 2.9 ghz, even while stress testing on p95, IBT at max, and so on. I've tried enabling/disabling speedstep, making sure turbo is definitely ON in the bios, no luck. The chip seems to always encounter throttling of some sort. I'm running the locked A05 bios, if that's of any insight.

cheers for the tips mate!

 

Did you enable the Extreme option in the BIOS and set a Dell Level 3 OC? If not, it will not run like an Extreme CPU is capable of running. It will run like a QM processor if the Extreme option is not enabled. Also, you might try disabling non-turbo flex support after enabling the Extreme option and setting a Dell Level 3 OC.

 

Me again lol so i have one more question i just recieved this replacement a week or so ago so dust isnt the issue. i was wondering if 90c is to hot for my 3920xm in crysis 2 and metro2033. it is a stable 90 it dont fluctuate or go higher according to hwinfo. are these ok or should i have dell repaste the Cpu. thanks in advance

 

That's hot. I'd call and ask for a technician to come out to check the cooling. Unless the ambient temp. In your home is high, it shouldn't exceed 70C.

Did you overclock it?

 

Are you serious??? 70C on the XM chip in Crysis? This would only be possible if your room is 10C, without any kind of overclock, with elevated back and a coolerpad. In normal conditions, without overclock, the CPU will still pull around 70W which would bring a 3-piped HS to its limits and produce 85-90C in demanding games. Also, remember, stock CPU fan profile is set to hit 100% rpm @85-88C, so those peak temps are pretty much inevitable.

 

Yeah, I would say hitting 90c with cpu intensive games, especially if overclocked, is pretty normal.

 

i have the level 3 overclock in the bios applied and the room i was playing in last night was around 78 degrees. so it looks like i should be ok i knew these xtreme chips run hotter just wasnt sure how hot. the gpus in game didnt get higher than 68

 

Looks like you are good to go. Nice gpu temps too. No need to repaste anything just yet :thumbsup:

 

Yes, I have to agree with Aikimox and Steve. Hitting 90°C on a machine that runs at 3.8GHz turbo by default is normal for a very demanding game like Crysis 2. As it is winter in Michigan and you may have the house warm because it is cold outside, it would be even more likely. If it does not go much higher and stay higher for long periods of time, I don't think you have any need for concern. The 3920XM T-junction is 105°C. If it gets that hot it will shut down to prevent any damage from taking place, so no reason for worry at this point.

If you have an unlocked BIOS you can lower the fan trigger point to go to 100% speed at a lower temperature, which will help lower the temps a little bit. If you don't have an unlocked BIOS, you can accomplish the same thing with HWiNFO64.

 

Mine never got up to that temp, so I assumed it was too hot. My bad.

Edit: Now that I think of it, it might have reached that temp. at one point. Maybe I was thinking of the GPU.

 

If you still have that 3940XM, you can easily test it.

 

I have a M18X R2 with a 3630qm cpu. Is 64 degrees a high idle temp for this system?

 

If the machine is really doing absolutely nothing, yeah that' not the correct idle temps. Of course, a lot depends on your ambient room temperature but I would say 64c is too hot for either gpu or cpu at idle.

 

Thanks for the reply! I figured as much. Guess it's time to whip out the ol' ic diamond.

 

My slot (0) GPU is running 5-10 degrees celsius cooler than my (1) gpu?

I thought that my 0 card was my primary, and my 1 card was my secondary. Isn't the secondary card suppose to run cooler than my primary? It just seems odd and backwards.

None of them ever get too hot to worry about but still.

 

Seems normal to me. I rarely saw both of them at the same temperature.

As long as they aren't near 90C, you're good.

 

Yes, that's normal.

Remember that the temps can be a reflection of more than load. If the left GPU heat sink and fan design, air flow/convection, etc is better than the right, that can easily explain the small difference. The other thing is, cooling the primary GPU is more important since the secondary usually runs cooler. So, from a bit of an odd point of view, it is actually good that you're seeing this instead of the alternative.

 

Okay I was just wanted to make sure. It just seemed odd that it was so much cooler when it had more load than the secondary. Thanks for your help guys

 

So what are people's temps running Crysis 3 ?
With a 4.0ghz - 4.1ghz OC I'm usually seeing things in the 80's with the very demanding scenes hitting low 90's for small but reasonable amounts of time.

 

Same here, Brother TBoneSan. Pretty much exactly the same.

 

Max cpu temps @ level 3 = 70c / max gpu temps stock clocks = 65c / Settings: All v.High/Blur Off/MSAAx2/2560x1600

Those were my test temps and I am yet to OC with Crysis 3 or mess around with game tweaks/settings.

 

Good stuff fellas :thumbsup: Things seem to look pretty cool for you there Brother Stevie, is that some of the Sandy Bridge magic you speak of ?

 

Well, I think the Sandy's run a tad cooler than Ivy does - I can't lay claim to any "magic" unfortunately!. The max temp I've seen on my cpu was 86c and that was during a 1024m ThrottleStop test with all cores running at 4.4Ghz. Also, don't recall seeing my gpu's anywhere above 65c yet - not OC'd them as of now. I guess that coupled with my NotePal U3 and the chilly British climate, it all adds up to some nice temperatures, PC wise. It's cool in my house too, probably not much more than 20c. It's also done on Dell stock Thermal Interface Material too, which goes to show that it's good stuff and highly effective.

 

I can confirm that. The Sandy Bridge shuts down at 100°C and tends to throttle more due to temps. Ivy Bridge does not reach that junction until 105°C, but it runs a little bit hotter at the same clock speeds.

 

Thanks for that, its good to know , Stevie's setup seems it runs very cool. I guess ambient temps might be the biggest X factor.
Its about 32F and less at nights here at the moment, but I have a air condition blowing warm air on me and and consequently in the machines general direction. I'm not raising the back or anything either. Hopefully I can keep similar temps in warmer weather.

 

After about an 45 mins of gaming @ 2540*1440 - Ghost Recon Future Soldiers

My temps are as follows :

CPU- Max 86 , idles between 45-55
GPU Temps - GPU 1 - Max 85
GPU2- Max 91 C

Dont you think the GPU temps are super high considering i did not overclock or do anything at all.

Also the CPU has the triple pipe heatsink.

 

Trooper, they do seem higher than normal. I don't think you should be seeing over 90c at all on your gpu's - that's a sign that something isn't quite right somewhere. Considering I run Crysis 3 (pretty much maxed) at stock gpu clocks and never see my cards go over 65c, 90c seems quite excessive. I am running at 2560x1600 by the way. The cpu seems a tad warm, especially if you have no OC on that either....