Solved

FX-6300 High Temp Concerns

Hello all! I've always used these forums as an excellent resource, and recently ran into a concern I thought the community could help with.

I recently pieced together a budget rig over the course of a few months snagging some parts on sale. For reference, this is what I'm working with:

AMD FX-6300 (stock settings)
Cooler Master Hyper 212 EVO (w/ Arctic Silver Arctic Alumina)
Gigabyte GA-970A-UD3P (stock settings)
Fractal Define R4 - additional 140mm fan added to front panel for 3 total; 2 in (front), 1 out (rear)

I had intentions to mildly overclock, though I have had no experience with it in the past. At the time of this post and the event to be detailed I have not changed any BIOS settings in an effort to overclock.

I completed putting the box together with my son over the weekend and proceeded to install Windows 8.1 (I had a key gifted to me, thought I'd see what all the fuss was about). I also booted with the case open to make sure all fans were turning. After which, I rebooted and entered BIOS, and noticed my CPU temperatures would jump from 27C to 34C fairly erratically. I then read some posts about AMD CPU temperature sensors being wonky at low readings sometimes and decided I should stress test it to both burn in and confirm that my temperatures were acceptable.

So Monday night I loaded Speedfan and Prime95 from a USB and set to work. First, it was difficult with Speedfan to tell what was what (I understand I should probably be using HWMonitor or something else, and probably will when I decide to move forward again). I found a post from another user with the exact same motherboard and processor with both Speedfan and HWMon pulled up, and it seemed to verify that Temp 2 was the socket temp and Temp 3 was the CPU temp.

At any rate, everything seemed fine. Idle was in the low 20s. Loaded for a few hours, the CPU seemed to hold around 32C and the socket around 40C. Since I was about to go to bed, I figured I would leave prime running overnight, just to make sure there were no errors or problems.

The following morning I was still around 33-34C on the CPU, maybe 41-42C on the socket. Glad that I was getting a consistent reading now, I headed to work. I was only supposed to work a half day, so I thought I'd be getting about 16 hours of prime in, and I'd move on with overclocking or what have you that afternoon.

Well, work had a catastrophe, and I ended up there until later than expected, and arrived home at about 20 hours of stress testing (I've tested for longer than this in the past and wasn't terribly concerned). After returning home I find that my Temp 2 and Temp 3 are 91C and 79C respectively! Alarmed, I immediately stopped the test (which had no errors or warnings), at which point the temps began to drop to around 50C.

After some investigation, I found that my CPU fan (included with the 212 EVO) had failed at some point. I then shut down the PC, restarting only a couple of times to verify that (a) the CPU fan did not work in any other fan header (it did not) and (b) that other known working fans did work in the CPU fan header (they did).

Luckily, I don't believe the CPU ran for too terribly long at this temperature because the mean temp for Temp 3 was still listed as 34.9C, max at 79C. (Speedfan's chart only goes back 15 minutes so I could not verify) Doing a little bit of raio math (which is at best a rough estimate), it seems the CPU ran like this for about one hour.

TL;DR: My FX-6300 ran a stress test for an unknown (but probably around an hour) amount of time with a Hyper 212 EVO heatsink but no CPU fan running to a maximum of 79C (I think, see questions).

I guess my concerns are multiple, but here they are:
Am I correct in my assumptions of the Temp 2 and Temp 3 being socket and CPU respectively?
Should I be concerned about the health of my CPU (or other hardware) after this incident?
Does this warrant me removing the heatsink and cleaning/reapplying thermal compound?
6 answers Last reply Best Answer
More about 6300 high temp concerns
  1. Well, its hard to say if it hurt the CPU or not. I mean AMD says safe temps are up to 62, so you were a fair amount over that for an hour. You most likely won't know until later on if the CPU has been damaged. I would keep using it and just see how it goes. It sounds like you have good airflow in the case so that helped you a bit.

    On my system, SpeedFan doesn't show CPU temps, only socket temps. I would definitely go with HWMonitor. That will show you all the temps and it is much easier to read. On HWMonitor, Temp1 is northbridge, temp2 is socket, and I'm not sure what temp3 is. But it has a separate spot telling you what CPU you have and tells temperature as well as voltage. It also has the same for your GPU. HWMonitor is free, so I would highly recommend downloading it and see what it says.

    Here is the link to download HWMonitor.

    http://www.cpuid.com/softwares/hwmonitor/versions-history.html

    Good luck!
  2. RookieOfTheYear said:
    Well, its hard to say if it hurt the CPU or not. I mean AMD says safe temps are up to 62, so you were a fair amount over that for an hour. You most likely won't know until later on if the CPU has been damaged. I would keep using it and just see how it goes. It sounds like you have good airflow in the case so that helped you a bit.

    On my system, SpeedFan doesn't show CPU temps, only socket temps. I would definitely go with HWMonitor. That will show you all the temps and it is much easier to read. On HWMonitor, Temp1 is northbridge, temp2 is socket, and I'm not sure what temp3 is. But it has a separate spot telling you what CPU you have and tells temperature as well as voltage. It also has the same for your GPU. HWMonitor is free, so I would highly recommend downloading it and see what it says.

    Here is the link to download HWMonitor.

    http://www.cpuid.com/softwares/hwmonitor/versions-history.html

    Good luck!


    Thanks for the feedback. I won't be able to fiddle with it again until I get home today (and maybe not even then - having twin young ones and a five year-old can be its own significant time constraint). My path forward for now I think will be getting a second 120mm fan while I wait on the warranty fan. Once I receive the replacement I will probably just add it to the heat sink in a push-pull configuration which also grants me a bit of redundancy if the same problem were to occur again. I will also get my hands on hwmon and get some initial readings with the new single fan installed.

    Looks like I won't be tinkering with overclocking anytime soon. I'd hate to add variables to the scenario.

    I will post my findings when I get them. Again, thanks for response!

    Just as a small followup:
    Have any of you seen an FX-6300 worked to death, and what were the circumstances?
    What should I set my CPU fans up for? PWM seems to be the right path, but what setting is good?
  3. So I had some time to play around with a cheap fan I could pick up from Best Buy, though it is obviously of lower quality than the one provided with the aftermarket cooler.

    I came prepared this time, with HWMon, Coretemp, and Speedfan. I noticed upon booting that the CPU temperature reading (along with the Temp3 from Speedfan which I had correlated with the CPU temperature) was erroneously low. I'm fairly certain something about the sensor was damaged here with the excessive temperature.

    I started up prime95 again just to see how everything would react. When the temperatures got above ~26C, they stabilized and stopped reading absurd numbers. I was close at hand to end the test if the temperatures did not resolve into something that made sense. I let the test run for about 10 minutes, wary of going beyond that with so many questions about the situation.

    The CPU temperature ran up to 40-41C (about 7 degrees higher than it averaged before the incident) and seemed to start to plateau there. I'm unsure of whether this could indicate a problem I've created with the excessive temperature, or be a result of the lower quality fan on the heatsink.

    I believe the Cooler Master fan was 87ish CFM wherreas this fan is 46ish max.

    In addition, when attempting to get the new fan on the Hyper 212 EVO heatsink, the heatsink shifted slightly over the top of the CPU. This alone is enough to make me want to clean and reapply thermal compound. The increased temperature at load simply adds to the concern.

    Any thoughts?
  4. Xcom46 said:
    If you would like a good amount of cooling.I have a suggestion here.

    I use all fans in my system and they keep my desktop pretty cool under heavy load.Fare warning be four hand the two main fans require a speed fan controller these items are pretty cheap and work better then that Speed fan crap.When looking into fans the two things to look for are CFM and RPM's

    If you are looking to upgrade later.

    LOGISYS Computer FP600BK 6-Channel Fan Controller
    http://www.newegg.com/Product/Product.aspx?Item=N82E16811998100

    This fan below is small but power full i have these sitting on my heat sink blowing out the hot air from the heat sink and the ram sticks.

    bgears b-Blaster 80 80mm Case Fan 3500 RPM 62 CFM
    http://www.newegg.com/Product/Product.aspx?Item=N82E16835132020

    These next fans are a half inch thick in width.And they are loud when running at 5200 rpms.I use one on the back panel to suck out the hot air quickly.And one on my side to blow on the GPU and PSU to keep them both cool.This is where the speed fan controller comes in.
    Delta AFB1212GHE-CF00 120mm Case cooler 5200 RPM 240.96 CFM
    http://www.newegg.com/Product/Product.aspx?Item=N82E16835213001

    When i bought them delta fans at the time they where like 54 dollars at the time.They dropped down to 26 dollars.

    This setup keeps my GPU at 35ºC = 95.000ºF normal work load gaming mode it stays at 40ºC = 104.00ºF

    And my AMD Phenom II x 4 965 (OC) to 4.20ghz keeps it at 30ºC = 86.000ºF normal load and under gaming mode stays at 38ºC = 100.40ºF

    And all so of coarse my desktop is in the basement where it stays cool all year round.


    Hope this helps you.



    I certainly appreciate the tips. However, my main concern here was the event I described in my first post. The CPU was worked at high temperatures due to a CPU fan failure, and was looking for advice on how to move forward with further testing.

    I will keep your points in mind, however, if I decide to add more cooling to my case. The current setup (when the CPU fan had not failed) kept my CPU at 33.5C when fully loaded.
  5. Best answer
    @xcom46 You should read the original post, he explains everything. He had a hyper 212 evo and the fan stopped working for about an hour before he was able to turn off the computer and temps soared to about 79. He is wondering how much me might have hurt the CPU.

    @devant0r...The FX6300 are notorious for sensors giving false readings. So hopefully that was all that was damaged when the CPU got that hot. I think you will be fine but be prepared for the CPU to one day start to fail or just stop all together. The good thing is that these CPU's are not very expensive. So you can replace it with a new one for only $120 if it does fail.

    I don't know if I would overclock it much past 4.2 Ghz. But then again who knows how strong your individual unit is (each CPU is a little different). I personally don't know any tricks or ways to find out the health of a CPU after an incident like this. Bu I would just be prepared for it to possibly fail. It might also last you a very long time, its just hard to tell.

    Good luck!
  6. RookieOfTheYear said:
    @xcom46 You should read the original post, he explains everything. He had a hyper 212 evo and the fan stopped working for about an hour before he was able to turn off the computer and temps soared to about 79. He is wondering how much me might have hurt the CPU.

    @devant0r...The FX6300 are notorious for sensors giving false readings. So hopefully that was all that was damaged when the CPU got that hot. I think you will be fine but be prepared for the CPU to one day start to fail or just stop all together. The good thing is that these CPU's are not very expensive. So you can replace it with a new one for only $120 if it does fail.

    I don't know if I would overclock it much past 4.2 Ghz. But then again who knows how strong your individual unit is (each CPU is a little different). I personally don't know any tricks or ways to find out the health of a CPU after an incident like this. Bu I would just be prepared for it to possibly fail. It might also last you a very long time, its just hard to tell.

    Good luck!


    Thanks for the responses.

    I got a new Cooler Master Blade Master 120mm fan in yesterday and installed it in place of the cheap Rocketfish I got at Best Buy. I played maybe four games of League of Legends (not the most resource intensive game I know) and peaked a temperature of 40C. I will probably try prime95 for an hour or so one night and see how it responds.

    I also got a PWM splitter, so I'll be putting the replacement 120mm warranty fan in push/pull with the Blademaster just for redundancy and added air flow.

    I guess for now it is just wait and see. I may try to overclock slightly, but it will be a while before I do so.

    Thanks again!
Ask a new question

Read More

Cooling CPUs AMD Temperature