• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

High CPU temp, should i return it?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
if you're only moving between ~93 C and ~96 C between idle and gaming, I'd say the sensor is wrong, you should warm up way more than 3 C. OEM laptops are sometimes funky to get the right values read. How warm does the bottom of the laptop feel (specifically in the corner where the fan is)?
 
ketchup79 said:
Hmm, strange. If you temps are really that high, that whole laptop should be quite hot all the time, as sm625 mentioned. Is it?
Click to expand...

not if the heat sink isnt function as it should be...

then it would take some time for the heat to transfer though the plastic shell for her to notice it.

Anyhow i have no clue on the trinity chip... im guessing the standard DTS sensors arent working properly, or the APU has a completely different set of DTS.
 
RampantAndroid said:
Can this CPU even hit 90C without throttling? Can these numbers even be accurate? 90C would be enough for it to feel VERY hot on the shell of the laptop. If my macbook hit 90 I'd likely burn myself on the metal case.
Click to expand...

Exactly, I am wondering if the programs just don't know how to read the sensor on that chip. If it were getting that hot, the thing would be throttling so crazy it would feel dog slow.
 
ketchup79 said:
Exactly, I am wondering if the programs just don't know how to read the sensor on that chip. If it were getting that hot, the thing would be throttling so crazy it would feel dog slow.
Click to expand...

yeah, its definitely not throttling or scolding hot or slow, that's for sure.
 
That Argus program looks like your best bet then. It was the only one showing you the temps for all the cores, and the numbers were realistic.
 
It sounds like you're reading the temp as 90F and not C. The other program that you read 33C sounds right. 33C = ~90F.
 
SarahK said:
That is a relief :'D what core temps are on the high side? Just for future reference.
Click to expand...

Laptops are tough. Temps will depend on the design, and the way they set the fan speed to ramp. Just for reference, the cores in the i7 in my laptop are running between 36 and 42 right now. (not running much of anything). As has been mentioned, your laptop will get quite warm, noisy, and the cpu will start to throttle to let you know if temps get too high.
 
ketchup79 said:
Laptops are tough. Temps will depend on the design, and the way they set the fan speed to ramp. Just for reference, the cores in the i7 in my laptop are running between 36 and 42 right now. (not running much of anything). As has been mentioned, your laptop will get quite warm, noisy, and the cpu will start to throttle to let you know if temps get too high.
Click to expand...

Alrighty. Thanks!
 
There's no way the computer would be operating normally if you had your chip running in the 90C range. That's throttle/thermal shutdown territory. I'd normally say your core sensor is borked, but if you're getting accurate readings from at least one program then your issue is almost certainly the monitoring software.

Temp monitoring software is weird stuff. Most manufacturers caution that it is at best a rough indication of actual temp and at worst downright wrong. Hell, even the people who created many of the programs have difficulty explaining why they simply don't work sometimes. You've just got to find the one that seems to be closest to the mark and stick with it.

Anyway, I'm quite confident that if you were loading over 90C you'd be having some very serious, very noticeable problems. I think you're good!
 
BigChickenJim said:
There's no way the computer would be operating normally if you had your chip running in the 90C range.
Click to expand...

Not at all. Current CPUs run perfectly fine at that temp. it's is just not advisable to do so due to power consumption and durability (especially when you OC).

While gaming heavy 3D game hitting 90°C on a laptop is IMHO pretty standard. Hell i ran BOINC on a crappy of the shelf PC with core2-duo E4300 for several years and load temps were easily 90°C and never an issue.

Laptops im general are just not made for prolonged high cpu load. they all get hot.
 
beginner99 said:
Not at all. Current CPUs run perfectly fine at that temp. it's is just not advisable to do so due to power consumption and durability (especially when you OC).

While gaming heavy 3D game hitting 90°C on a laptop is IMHO pretty standard. Hell i ran BOINC on a crappy of the shelf PC with core2-duo E4300 for several years and load temps were easily 90°C and never an issue.

Laptops im general are just not made for prolonged high cpu load. they all get hot.
Click to expand...

Sorry, don't buy it. 90C is not a safe idle (or load, for that matter) temperature and it most certainly isn't a "standard" temp even in a laptop. The OP's chip has a max operating temp of 100C. It's going to throttle well before it reaches that, and if it were doing that she (assuming OP is a she) would have definitely noticed.

I gamed heavily on a laptop for years and never saw temps over 65C. The fact that you ran your entirely different hardware dangerously hot and got lucky doesn't mean it's normal or advisable to do so.

Regardless, the OP is receiving sensible readings from alternate monitoring software and there are multiple accounts out there of this processor not polling correctly in some programs. The issue is software-related.
 
Last edited:
Pretty easy to figure out if its a wrong sensor reading... when the laptop is idle does the fan speed slow down? And then speed up when putting a load on the CPU? Why I say this is because the fan curve would be on max even at 90c. So the fan would run at 100% 24/7 if the sensor was correct. So if your fan is throttling while you game and quiet on the desktop then you should be good to go. The fan curve normally don't start until 60c on a laptop and by the time it hits 85c the fan should be around max rpm.
 
Last edited:
BigChickenJim said:
......

Temp monitoring software is weird stuff. Most manufacturers caution that it is at best a rough indication of actual temp and at worst downright wrong. Hell, even the people who created many of the programs have difficulty explaining why they simply don't work sometimes. You've just got to find the one that seems to be closest to the mark and stick with it.
.....
Click to expand...

I think its just AMD that is weird stuff. Intel cpus don't seem to have problems reporting actual temps instead of some arbitrary scale like AMD.
I don't remember Intel cautioning that their reported temps might be downright wrong.
 
bononos said:
I think its just AMD that is weird stuff. Intel cpus don't seem to have problems reporting actual temps instead of some arbitrary scale like AMD.
I don't remember Intel cautioning that their reported temps might be downright wrong.
Click to expand...

Are you really trying to turn this into an AMD vs. Intel argument? No, not going to bite.

Third-party monitoring software is often inaccurate regardless of manufacturer. Anyone who has ever relied on it knows this. That's all there is to it.
 
BigChickenJim said:
Are you really trying to turn this into an AMD vs. Intel argument? No, not going to bite.

Third-party monitoring software is often inaccurate regardless of manufacturer. Anyone who has ever relied on it knows this. That's all there is to it.
Click to expand...

I have used different monitoring apps on my last 5 Intel systems, and they were all pretty accurate (all showed the same temps on same system, seemed reasonable numbers) so their could be something to bononos' statement.
 
BigChickenJim said:
Are you really trying to turn this into an AMD vs. Intel argument? No, not going to bite.

Third-party monitoring software is often inaccurate regardless of manufacturer. Anyone who has ever relied on it knows this. That's all there is to it.
Click to expand...

No, just stating something thats really obvious which has been going on for several cpu generations.
And no, I don't remember Intel saying in any of its whitepapers that the temps might be 'downright wrong'. AMD says that whatever is being reported are not temps but some figure on a non-linear scale.
 
That's true as mentioned in my link before: "The value for the CPU temperature of the (digital) on-chip sensor does not equal any real physical temperature but is measured on an AMD spezific scale. This was done to be able to provide more accurate values when the temperature is close to the thermal design limit of the processor and is used to throttle the CPU to prevent damage by overheating. Therefore reliable measurements for CPU idle states is not possible."

But this wouldn't explain the 90c readings, I do remember when using coretemp with my old Athlonx2x64 I had top go into the options to select a different temperature sensor to get more accurate temps but even then the Brisbane core was notoriously innacurate when it came to temps.

Still if Argus monitor shows 33c and the Laptop isn't hot then I'd say the OP has nothing to worry about.
 
bononos said:
No, just stating something thats really obvious which has been going on for several cpu generations.
And no, I don't remember Intel saying in any of its whitepapers that the temps might be 'downright wrong'. AMD says that whatever is being reported are not temps but some figure on a non-linear scale.
Click to expand...

I'll just leave this here.

http://www.intel.com/support/processors/sb/CS-033342.htm

"Most third-party applications are reporting the processor temperature fairly accurately. Because reports can sometimes show erroneous data, download Intel® Processor diagnostic Tool to test if your processor is working correctly."

Fairly accurately = rough indication.
Erroneous = downright wrong.

I'm not saying that Intel isn't more accurate with temp reporting than AMD, but to try to brush the issue off entirely and act as if any problems are AMD's fault is misleading. I'm sticking with my original statement: third-party temp monitoring is sometimes subject to strange behavior and incorrect readings no matter which manufacturer made your components.
 
Last edited:
BigChickenJim said:
I'll just leave this here.

http://www.intel.com/support/processors/sb/CS-033342.htm

"Most third-party applications are reporting the processor temperature fairly accurately. Because reports can sometimes show erroneous data, download Intel® Processor diagnostic Tool to test if your processor is working correctly."

Fairly accurately = rough indication.
Erroneous = downright wrong.

I'm not saying that Intel isn't more accurate with temp reporting than AMD, but to try to brush the issue off entirely and act as if any problems are AMD's fault is misleading. I'm sticking with my original statement: third-party temp monitoring is sometimes subject to strange behavior and incorrect readings no matter which manufacturer made your components.
Click to expand...

Intel is being conservative in the doc you linked. The temps reported from recent Intel generations are a little better than just a "rough indication". And any erroneous reports are most probably very infrequent momentary blips having never encountered something like it myself and definitely not "often inaccurate" as you said in your earlier post.

AMD on the other hand, has a common problem with lower than ambient temps, and even -ve temps from the older Athlons up till the Phenoms. From the Phenoms on, AMD doesn't report actual temps but something else with an offset - good luck converting those to temps.
The Bulldozer/PD cpus doesn't have the thermistors on the actual cores but they are somewhere else on the package and the reported core temps are actually extrapolated from those (unlike Intel where the diode is surrounded by the cores or somewhere close and the temps actually correspond to where the cores sit in relation to the overall die) making BD/PD reporting less accurate or way less accurate if the backup Tjunction diode is used for reporting.
 
Last edited:
You must log in or register to reply here.

TRENDING THREADS

Back
Top