Sign-in / Sign-up
Your question

2600K temperature hot with CM Hyper 212+

Tags:
  • Heatsinks
  • Cooler Master
  • Temperature
  • Overclocking
  • Product
Last response: in Overclocking
September 25, 2012 2:54:09 PM

Hey, guys.


I am OC newbie who needs expert advises. I really need help with troubleshooting 2600K temperature (OC'd or Stock speed), using Cooler Master Hyper+ with AS5. I have tried OC @ 4.4Ghz, and also stock @ 3.4Ghz, and both temperatures seem high.

I have been doing extensive research on what reasonable temperature for 2600K should be around 65°C @4.5ghz, and for some reason, my rig's temperature seem to be always 10~15°C+ higher than most of other similar setup. No matter what I do, the temperature will not decrease enough to justify a "reasonable" thermal setup. I need expert advises on what I could do next, or I may consider purchasing new mobo + cpu, because this issue is sort of like a splinter in my finger (I know it won't kill me, but it is driving me nuts). PLEASE HELPPP!!!

Test Results:
• With BIOS setting to auto:
• CPU @4.4Ghz max temp was at 78~81°C at full load “blend”
• @4.0ghz max temp was at 74~75°C at full load “blend”
• @4.0ghz, HT OFF, max temp was at 70°C at full load “blend”
• @stock 3.4ghz, with bios settings set to Auto, max temp was at 65°C at full load “blend”

Additional Test Info:
• Ambient Temp was always 26~27°C.
• Idle Temp was around 33~37°C.
• 3rd core was always the hottest, 1st core always the coolest. Difference was around 2~9°C, depending on load.
• Temp was measured with HW Monitor & Speedfan.
• Full load tested with Prime95 “blend” and/or “large FFT”
• Test conducted for min of 20~30 min. temps would rise to near the max temp within the first 30min. When testing overnight, the temps would rise additional 2° ish. Example: @ 4.4Ghz full load, temp would rise to around 69~70° in 10 min, and to around 76~78° the next 10 min. So within 20 min, it nearly reaches the max temp.
• For some reason, Prime95 “large FFT” on some tests had lower max temp than “blend,” maxing out at around 75°. (Unsure of stability though)
• CPU-Z core volt is around 1.28 ~ 1.33 volt. The fluctuation varies somewhat from AUTO to manual voltage set in Bios, but it is usually under 1.3volt at full load shown here.

What I tried:
• I tested with Xigmatek DK S1283V rev.w (NOT SD1283) and changed to CM Hyper+ to compare results. Not much difference. May be 1~2°C max.
• I tried to apply AS5 using different application methods (size of bb in middle, spread, line method), but the temp difference was minimal, usually about 2~3°C max. Best result seem to be spread method. Questioning the shelf life of 5yr old AS5, purchased “new” AS5, and reapplied. No difference.
• Changed Bios Setting from Auto to manual core voltage at 1.31 (the lowest stable volt). Played around with other voltages and minor tweaks, but no difference.
• Changed Bios core volt to 1.29, and played around with difference LLC levels, but still the max temp difference isn’t that much. And PC is difficult to stabilize.
• I tried several different fan setup (2 fans, 3 fans, 4 fans with different combination of side intake, side exhause, top exhause, etc) as well as push and pull with Hyper+, but no major changes. May be 2~3°C difference.
• With the case side panel off, it would decrease the temp by 2~4°C max. Not TOO difference from different fan setup.
• Tried Gigabyte Easy OC Tool. No difference.


Hardware Info:
Case: CM Elite 430 Case
PSU: OCZ ZS Series 650W OCZ-ZS650W
CPU: i7-2600K
Mobo: GA-Z68X-UD3H-B3
HS: CM Hyper 212+ / Xigmatek Dark Knight-S1283V REV.W
RAM: Patriot Viper 3 16GB (4 x 4GB) DDR3 1600
GPU: XFX Radeon HD 6850 1GB (HD-685X-ZNFE VC.4)
SSD: OCZ Vertex 3 90GB SSD x2 raid0
HDD: SAMSUNG EcoGreen F4 HD204UI 2TB, WD Caviar Blue WD5000AAKX 500GB
ODD: LITE-ON DVD Burner iHAS124-04
OS: Win7 SP1 64bit

More about : 2600k temperature hot hyper 212

September 25, 2012 3:22:12 PM

Where are you located, that ambient temperature is on the high side?
m
0
l
September 25, 2012 3:24:48 PM

Jim_L9 said:
Where are you located, that ambient temperature is on the high side?

I live in Texas. Even with the high ambient temp, I think hitting nearly 80C on OC to 4.4ghz is high. I tried to OC to 4.5ghz, and my temps jumped up to over 82C...
m
0
l
Related resources
September 25, 2012 3:48:56 PM

I'd recommend a self contained water cooler for your CPU since it sounds like the ambient temps are the problem. Or you could crank your A/C all the way down and see if that helps.
m
0
l
September 25, 2012 4:11:24 PM

MotherFerJones said:
I'd recommend a self contained water cooler for your CPU since it sounds like the ambient temps are the problem. Or you could crank your A/C all the way down and see if that helps.


I thought ambient temp increases/decrease the core temp 1:1 ratio? What I am asking is if I drop the ambient temp 6C (26C to 20C), then doesn't the max temp would drop 6C as well (80C to 74C)? Or is that logic flawed?
m
0
l
September 26, 2012 5:04:45 AM

sounds like an experiment to me, dont blame me when you get the electric bill though lol. In theory 1:1 is right but the heatsink is not going to be 100% effecient so its probably more like 1:.5-.75 depending on the efficiency of the heatsink/ fan combo.
m
0
l
a c 152 K Overclocking
September 26, 2012 5:09:20 AM

MotherFerJones said:
I'd recommend a self contained water cooler for your CPU since it sounds like the ambient temps are the problem. Or you could crank your A/C all the way down and see if that helps.

NO. :pfff: 
m
0
l
a c 152 K Overclocking
September 26, 2012 5:10:33 AM

How is case airflow within the 430? It's a fairly small case that offers minimal airflow. Did you clear up cables?

BTW 80's is perfectly normal on a 212. My 2600K @4.5ghz 1.31V was hitting the mid 80's under Prime95.
m
0
l
September 26, 2012 5:37:24 AM

Isn't the only difference between an i5 2500k and a i7 2600k hyperthreading?

60C max tempurature even with an hour of intel burn test on a 4.5ghz i5 2500k.
Im thinking maybe the thermal paste was added wrong or it hasnt been long since you added. I think i read somewhere that it takes a couple hours of use for thermal paste to be working at its best.
m
0
l
a c 152 K Overclocking
September 26, 2012 6:02:08 AM

No, they run at different voltages.
m
0
l
September 26, 2012 1:30:30 PM

amuffin said:
How is case airflow within the 430? It's a fairly small case that offers minimal airflow. Did you clear up cables?

BTW 80's is perfectly normal on a 212. My 2600K @4.5ghz 1.31V was hitting the mid 80's under Prime95.


Thank you for your input. Your feedback makes me feel better. I simply thought Hyper212+ had better results for pretty much everyone else, and assumed my rig wasn't right.

Well, I did my best to clear all cables, but CM430 case doesn't offer much space to put them away. I guess I could try to organize them differently, but I thought that my current temp were still too high, and airflow inside the case didn't have that much impact.

I re-ran Prime95 "blend" mode after re-applying AS5 (this time, used spread method, on both heatsink + cpu). Fan is still the default setup (4 fans: 1 front in, 1 hyper212 push, 1 rear ex, 1 side ex), ambient at 26C, Bios at Auto, and CPU-Z core volt hitting 1.296v at full load.

Test1: 3 fans, side panel off, no videocard - max temp 74~75, after 20 min.
Test2: 4 fans, no videocard - max temp 75, after 20 min.
Test3: 4 fans, w/videocard - max temp 76, after 20 min.

So like I said before, spread method seem to give me the best results. Any ideas on what I could try next to trim off a few more degrees?
m
0
l
September 26, 2012 1:44:33 PM

footlong96 said:
Isn't the only difference between an i5 2500k and a i7 2600k hyperthreading?

60C max tempurature even with an hour of intel burn test on a 4.5ghz i5 2500k.
Im thinking maybe the thermal paste was added wrong or it hasnt been long since you added. I think i read somewhere that it takes a couple hours of use for thermal paste to be working at its best.


Sounds like a typical max temp reading I have been hearing about: @4.5ghz, 2500K about 60C max, and 2600K about 65~67C max, under typical conditions, typical setup. My temps are still about 10C higher than "typical" temps. If I could trim off 5~6C, I would be happy and DONE! Disabling HT does drop the temp about 5C, but I don't plan to do that, since it defeat the purpose of buying 2600K and disable HT.
m
0
l
a c 152 K Overclocking
September 26, 2012 2:20:24 PM

Those are custom water cooling temps. :) 
m
0
l
a b K Overclocking
September 26, 2012 2:43:49 PM

goonbar79 said:
Hey, guys.


I am OC newbie who needs expert advises. I really need help with troubleshooting 2600K temperature (OC'd or Stock speed), using Cooler Master Hyper+ with AS5. I have tried OC @ 4.4Ghz, and also stock @ 3.4Ghz, and both temperatures seem high.

I have been doing extensive research on what reasonable temperature for 2600K should be around 65°C @4.5ghz, and for some reason, my rig's temperature seem to be always 10~15°C+ higher than most of other similar setup. No matter what I do, the temperature will not decrease enough to justify a "reasonable" thermal setup. I need expert advises on what I could do next, or I may consider purchasing new mobo + cpu, because this issue is sort of like a splinter in my finger (I know it won't kill me, but it is driving me nuts). PLEASE HELPPP!!!

Test Results:
• With BIOS setting to auto:
• CPU @4.4Ghz max temp was at 78~81°C at full load “blend”
• @4.0ghz max temp was at 74~75°C at full load “blend”
• @4.0ghz, HT OFF, max temp was at 70°C at full load “blend”
• @stock 3.4ghz, with bios settings set to Auto, max temp was at 65°C at full load “blend”

Additional Test Info:
• Ambient Temp was always 26~27°C.
• Idle Temp was around 33~37°C.
• 3rd core was always the hottest, 1st core always the coolest. Difference was around 2~9°C, depending on load.
• Temp was measured with HW Monitor & Speedfan.
• Full load tested with Prime95 “blend” and/or “large FFT”
• Test conducted for min of 20~30 min. temps would rise to near the max temp within the first 30min. When testing overnight, the temps would rise additional 2° ish. Example: @ 4.4Ghz full load, temp would rise to around 69~70° in 10 min, and to around 76~78° the next 10 min. So within 20 min, it nearly reaches the max temp.
• For some reason, Prime95 “large FFT” on some tests had lower max temp than “blend,” maxing out at around 75°. (Unsure of stability though)
• CPU-Z core volt is around 1.28 ~ 1.33 volt. The fluctuation varies somewhat from AUTO to manual voltage set in Bios, but it is usually under 1.3volt at full load shown here.

What I tried:
• I tested with Xigmatek DK S1283V rev.w (NOT SD1283) and changed to CM Hyper+ to compare results. Not much difference. May be 1~2°C max.
• I tried to apply AS5 using different application methods (size of bb in middle, spread, line method), but the temp difference was minimal, usually about 2~3°C max. Best result seem to be spread method. Questioning the shelf life of 5yr old AS5, purchased “new” AS5, and reapplied. No difference.
• Changed Bios Setting from Auto to manual core voltage at 1.31 (the lowest stable volt). Played around with other voltages and minor tweaks, but no difference.
• Changed Bios core volt to 1.29, and played around with difference LLC levels, but still the max temp difference isn’t that much. And PC is difficult to stabilize.
• I tried several different fan setup (2 fans, 3 fans, 4 fans with different combination of side intake, side exhause, top exhause, etc) as well as push and pull with Hyper+, but no major changes. May be 2~3°C difference.
• With the case side panel off, it would decrease the temp by 2~4°C max. Not TOO difference from different fan setup.
• Tried Gigabyte Easy OC Tool. No difference.


Hardware Info:
Case: CM Elite 430 Case
PSU: OCZ ZS Series 650W OCZ-ZS650W
CPU: i7-2600K
Mobo: GA-Z68X-UD3H-B3
HS: CM Hyper 212+ / Xigmatek Dark Knight-S1283V REV.W
RAM: Patriot Viper 3 16GB (4 x 4GB) DDR3 1600
GPU: XFX Radeon HD 6850 1GB (HD-685X-ZNFE VC.4)
SSD: OCZ Vertex 3 90GB SSD x2 raid0
HDD: SAMSUNG EcoGreen F4 HD204UI 2TB, WD Caviar Blue WD5000AAKX 500GB
ODD: LITE-ON DVD Burner iHAS124-04
OS: Win7 SP1 64bit


From the looks of your case I sure hope you added more cooling fans then the stock 1. If you can't get the hot air out it will never cool well.
m
0
l
September 26, 2012 4:36:56 PM

ahnilated said:
From the looks of your case I sure hope you added more cooling fans then the stock 1. If you can't get the hot air out it will never cool well.


One questions about this. If I leave the side panel off, doesn't that result in lowest possible max core temp? Would excellent fan setup be better than no side panel? As you can see from what I have tried, leaving side panel off improves temp by a few degrees only.
m
0
l
September 26, 2012 4:52:44 PM

amuffin said:
Those are custom water cooling temps. :) 


Actually, these are air-cooling temps using hyper212+, or other similarly priced heatsinks (hyper212 evo, loki, etc). I've read through tons of hours of info on newegg reviews, various OCing forums, product reviews for 2600K, many heatsinks, my mobo, case airflow articles, thermal paste, etc, and that is the general info, taking into acct ambient temp and voltages. An example is overclockersclub.com, which shows that max temp wise Loki is 50C over, Dark Knight SD1283 is 42C over, Evo is 44C over ambient, using STOCK thermal paste.

At this point, I will be happy with just a few degrees drop using any method necessary, without disabling features.

m
0
l
a b K Overclocking
September 26, 2012 5:08:55 PM

goonbar79 said:
One questions about this. If I leave the side panel off, doesn't that result in lowest possible max core temp? Would excellent fan setup be better than no side panel? As you can see from what I have tried, leaving side panel off improves temp by a few degrees only.


In some cases yes, in others no. It depends on how the air flows through the case and if it can be evacuated.
m
0
l
September 27, 2012 1:32:53 AM

if your super hard up for a tad extra cooling you could hook the CPU heatsink fan up to a three pin mobo connector so it runs full speed all the time, but if you cant get all that hot air out of the case efficiently then it wont help much.
m
0
l
a c 152 K Overclocking
September 27, 2012 1:40:42 AM

goonbar79 said:
Actually, these are air-cooling temps using hyper212+, or other similarly priced heatsinks (hyper212 evo, loki, etc). I've read through tons of hours of info on newegg reviews, various OCing forums, product reviews for 2600K, many heatsinks, my mobo, case airflow articles, thermal paste, etc, and that is the general info, taking into acct ambient temp and voltages. An example is overclockersclub.com, which shows that max temp wise Loki is 50C over, Dark Knight SD1283 is 42C over, Evo is 44C over ambient, using STOCK thermal paste.

At this point, I will be happy with just a few degrees drop using any method necessary, without disabling features.

Ah, I didn't see you say "typical" temps. As well as now you're saying the temps over ambient.
Custom Loop gives you much lower temps. Around 50's with IBT and Prime95, as well as 5-7C within ambient temps during idling and 10-15C during normal use.
m
0
l
September 27, 2012 6:58:00 AM

ahnilated said:
From the looks of your case I sure hope you added more cooling fans then the stock 1. If you can't get the hot air out it will never cool well.


^+1
About a week ago i bought three fans for my case. I already had one exhaust but no intakes. added two intakes and another exhaust and it drastically helped out in cooling my cpu and graphics cards. My graphics cards are right under my cpu so naturally heat rises and goes to my cpu fan pushing hotter air into my evos fins. One i added my intake fan (which on my case is on the side right above my cpu heatsink) my cpu dropped a good 10C. Used to be a lot hotter with only one stock fan in there :p 
m
0
l
September 28, 2012 1:06:25 PM

footlong96 said:
^+1
About a week ago i bought three fans for my case. I already had one exhaust but no intakes. added two intakes and another exhaust and it drastically helped out in cooling my cpu and graphics cards. My graphics cards are right under my cpu so naturally heat rises and goes to my cpu fan pushing hotter air into my evos fins. One i added my intake fan (which on my case is on the side right above my cpu heatsink) my cpu dropped a good 10C. Used to be a lot hotter with only one stock fan in there :p 


Okay, so after reading the advises about the airflow inside my case, I went back and spend 1~2 hrs to re-do the cabling so most of the cables are nicely tied under the ODD / HDD trays, and virtually no cables tied up around the heatsink, videocard, etc. Also, I went back and changed the fan setup a few times, all 120mm size. Here are the results:

BEFORE RE-CABLING
Base setup: 1 front intake, 1 rear exhaust, 1 side exhaust, 1 heatsink fan (mounted behind heatsink as exhaust) = 76C max.
new setup 2: 1 front intake, 2 top intake, 1 side exhause, 1 rear exhaust, 1 heatsink fan = 77C max

AFTER RE-CABLING
Base setup: 1 front intake, 1 rear exhaust, 1 side exhaust, 1 heatsink fan (mounted behind heatsink) = 75~76C max.
new setup 1: 1 front intake, 1 side intake, 1 rear exhaust, 1 top exhaust, 1 heatsink fan = 75~76C max
new setup 2: 1 front intake, 1 side intake, 1 rear exhaust, 1 top exhaust, 1 heatsink fan = 75~76C max

I guess airflow isn't the core issue.. arrrrgggg
m
0
l
September 30, 2012 1:41:11 AM

The top of your case should be exhaust not intake. heat rises yo.
m
0
l
October 1, 2012 12:31:28 PM

footlong96 said:
The top of your case should be exhaust not intake. heat rises yo.


You can see that most of the setup I tried are with top exhaust. Since for my build, side exhaust setup wasn't too bad, I just tried stronger top intake with side exhaust to see the difference for comparison. It is just frustrating that no fan setup changes enough temperature within my case.

I am starting to be fairly certain that it may be my cpu and/or mobo combo simply generating more heat, because:

1) I tried 2 different heatsink with similar results
2) the cpu is hot under load at even at stock speed
3) no fan setup changes much temp.

And I don't think the ambient temp is the reason, because 25~26C should be fairly typical for a general PC environment.
m
0
l