Overclock.net › Forums › AMD › AMD CPUs › ★ ˜1090t Overclocking/Temperature Confusion ★
New Posts  All Forums:Forum Nav:

★ ˜1090t Overclocking/Temperature Confusion ★

post #1 of 40
Thread Starter 
Trying to get my 1090t to a 4GHz+ OC on my Asus M4A89GTD (BIOS 2301). I'm cooling it with a Corsair H100 in a HAFX mounted on top as an exhaust in a push configuration (on performance profile/max rad fan speeds). All other case fans are stock. I am just confused about which temps I am supposed to be reading.

Currently I am at 4.063GHz, which survived 6 or 7 consecutive runs of Cinebench 11.5 CPU, as well as ~20 minutes of Prime95 (before I stopped it @64°C).
BIOS Settings:
13.5x multiplier
300 bus
ram @1600
cpu/nb is on auto
cpu voltage is set @1.415625v ->cpu-z, HWmonitor, and Probe II all report no lower than 1.428 idle, and as high as 1.476 @load.
both cpu LLC and cpu/nb LLC are ON
pcie and cpu spread spectrum are both OFF
c1e OFF
cool'n'quiet OFF
asus core unlocker OFF

Ambient temps are ~20°C.
HWMonitor, Asus PC Probe II and the BIOS are all reporting idle CPU temps of 38-40°C, and I hit 64°C at load in Prime95.
Both Core Temp and HWMonitor are reporting the core temps to be around 27 or 28°C at idle, 50°C at load.

Idle temps of 40°C seem unusually high for an h100, which leads me to believe the core temps are the ones I should be watching. I have done plenty of googling and found lots of people asking this same question, but the answers are always different. Some people say to watch the core temp because it is using the diode in the actual chip, and the CPU temp is reported by the diode on the mobo/socket. I also read others say that CPU temp is the one to be watching. I should also note that while most of the time they are all the same, HWMonitor is actually reporting different core temps for each core, even though this chip is supposed to only have one diode. Hopefully someone can help me figure this out.
post #2 of 40
No, with Thuban the core temperature sensors are always wrong by 8-12C. The CPU "socket/junction" temperature is the one that matters, but yes- those are roughly the same temperatures I get with an H80 @ 4.2Ghz, which does seem a little high with your H100.

Have you tried reseating the cooler? Which thermal paste are you using?

EDIT: Just noticed you have a push-only configuration. The H100 uses a thinner radiator design than the H80, so cutting the fan density would definitely hamper it in my opinion. Is it at possible for you to try push-pull?
Edited by mezmenir - 4/20/12 at 8:02pm
    
CPUMotherboardGraphicsGraphics
Intel Core i7-4790K ASRock Fatal1ty Z97X NVIDIA GeForce GTX 780 NVIDIA GeForce GTX 780 
RAMHard DriveHard DriveHard Drive
32GB ADATA XPG DDR3-2400 1x 256GB Plextor M6e M.2-PCIe 2x 256GB Samsung SSD 830 2x 2TB Western Digital Black 
OSMonitorKeyboardPower
Microsoft Windows 6.4.9841 CTP Qnix QX2710LED Logitech G710 EVGA Supernova 1000 G2 
CaseMouseAudio
Corsair Carbide 500R Logitech G3 Creative Zx, Sennheiser HD598 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel Core i7-4790K ASRock Fatal1ty Z97X NVIDIA GeForce GTX 780 NVIDIA GeForce GTX 780 
RAMHard DriveHard DriveHard Drive
32GB ADATA XPG DDR3-2400 1x 256GB Plextor M6e M.2-PCIe 2x 256GB Samsung SSD 830 2x 2TB Western Digital Black 
OSMonitorKeyboardPower
Microsoft Windows 6.4.9841 CTP Qnix QX2710LED Logitech G710 EVGA Supernova 1000 G2 
CaseMouseAudio
Corsair Carbide 500R Logitech G3 Creative Zx, Sennheiser HD598 
  hide details  
Reply
post #3 of 40
Thread Starter 
Using the stock thermal paste that came on it. I have not tried re-seating, should I get some more thermal paste before re-seating (probably a stupid question)? This was the first time I had ever installed a heatsink, so it's definitely possible I didn't seat it properly. I have however checked the screws and they are done up beyond finger tight still so I don't think that would be the problem.
post #4 of 40
It's not only the screws being tight that's important, make sure the block is aligned on the CPU right. I noticed that problem with my H80 pretty bad, where tightening the screws would twist the pump/block a little and screw up my thermal paste. Also make sure you tighten them in an interleaved fashion. Don't just crank one down, and then the other. That does terrible things to your thermal paste application.

The stock paste on the CWC80/100 used to be really good but if I remember right, it isn't the same as it used to be. I remember seeing debates about this.

Personally I recommend Arctic Silver 5 thermal paste if you try a reapplication of the block, which you will NEED new thermal paste for. And high-purity Alcohol for cleaning off the old goo.

I know there -are- higher performance pastes, but AS5 has always worked great for me, and tons of other folks, and the cost ain't that bad thumb.gif

AS5: http://www.newegg.com/Product/Product.aspx?Item=N82E16835100007

Application guides can be found in many places on this site, but as always- grain of rice in the middle of the CPU works well.

Edited for censoring the syn' to "twist".
Edited by mezmenir - 4/20/12 at 8:15pm
    
CPUMotherboardGraphicsGraphics
Intel Core i7-4790K ASRock Fatal1ty Z97X NVIDIA GeForce GTX 780 NVIDIA GeForce GTX 780 
RAMHard DriveHard DriveHard Drive
32GB ADATA XPG DDR3-2400 1x 256GB Plextor M6e M.2-PCIe 2x 256GB Samsung SSD 830 2x 2TB Western Digital Black 
OSMonitorKeyboardPower
Microsoft Windows 6.4.9841 CTP Qnix QX2710LED Logitech G710 EVGA Supernova 1000 G2 
CaseMouseAudio
Corsair Carbide 500R Logitech G3 Creative Zx, Sennheiser HD598 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel Core i7-4790K ASRock Fatal1ty Z97X NVIDIA GeForce GTX 780 NVIDIA GeForce GTX 780 
RAMHard DriveHard DriveHard Drive
32GB ADATA XPG DDR3-2400 1x 256GB Plextor M6e M.2-PCIe 2x 256GB Samsung SSD 830 2x 2TB Western Digital Black 
OSMonitorKeyboardPower
Microsoft Windows 6.4.9841 CTP Qnix QX2710LED Logitech G710 EVGA Supernova 1000 G2 
CaseMouseAudio
Corsair Carbide 500R Logitech G3 Creative Zx, Sennheiser HD598 
  hide details  
Reply
post #5 of 40
Thread Starter 
I guess I will pick up some artic silver 5, and try re-seating. I was considering buying two more fans and doing push/pull but I couldn't decide if it would be worth it. I was looking at cougar vortexes.
post #6 of 40
I have seen good things about the Cougar fans, and push-pull is definitely worth it, especially if you can shroud the fans. Planning on doing that myself when my Carbide case gets here Monday, it should have enough room for a shroud on the radiator 'push' fan.

By shroud, I mean using a fan frame to put distance between the fan and radiator, a "tunnel" if you will. It reduces the dead surface area of the radiator caused by the fan's motor/hub.
    
CPUMotherboardGraphicsGraphics
Intel Core i7-4790K ASRock Fatal1ty Z97X NVIDIA GeForce GTX 780 NVIDIA GeForce GTX 780 
RAMHard DriveHard DriveHard Drive
32GB ADATA XPG DDR3-2400 1x 256GB Plextor M6e M.2-PCIe 2x 256GB Samsung SSD 830 2x 2TB Western Digital Black 
OSMonitorKeyboardPower
Microsoft Windows 6.4.9841 CTP Qnix QX2710LED Logitech G710 EVGA Supernova 1000 G2 
CaseMouseAudio
Corsair Carbide 500R Logitech G3 Creative Zx, Sennheiser HD598 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel Core i7-4790K ASRock Fatal1ty Z97X NVIDIA GeForce GTX 780 NVIDIA GeForce GTX 780 
RAMHard DriveHard DriveHard Drive
32GB ADATA XPG DDR3-2400 1x 256GB Plextor M6e M.2-PCIe 2x 256GB Samsung SSD 830 2x 2TB Western Digital Black 
OSMonitorKeyboardPower
Microsoft Windows 6.4.9841 CTP Qnix QX2710LED Logitech G710 EVGA Supernova 1000 G2 
CaseMouseAudio
Corsair Carbide 500R Logitech G3 Creative Zx, Sennheiser HD598 
  hide details  
Reply
post #7 of 40
Thread Starter 
Hmm never heard of shrouding fans. I will have to see if I can find something to use as a spacer since I have no dead fans to use
post #8 of 40
Quote:
Originally Posted by alduin View Post

I guess I will pick up some artic silver 5, and try re-seating. I was considering buying two more fans and doing push/pull but I couldn't decide if it would be worth it. I was looking at cougar vortexes.

I wouldn't suggest that. Corsair switched from Shin-Etsu to Dow-Corning - both are at worst equal to Arctic Silver 5. I use the stock TIM on my H80 and right now my idle temps are 27 core/ 35 degrees CPU sensor. I should note that Thubans read temps inaccurately at idle, but that's another story.

I should also note that I use Arctic 5 in my other build, so I'm not badmouthing it in any way. smile.gif
Quote:
Originally Posted by mezmenir View Post

It's not only the screws being tight that's important, make sure the block is aligned on the CPU right. I noticed that problem with my H80 pretty bad, where tightening the screws would twist the pump/block a little and screw up my thermal paste. Also make sure you tighten them in an interleaved fashion. Don't just crank one down, and then the other. That does terrible things to your thermal paste application.

He's 100% spot on with that part. 9 times out of 10 thermal application is your cause for high temperatures.
    
CPUMotherboardGraphicsRAM
AMD Phenom X6 1090T ASUS M5A97 EVO Sapphire HD 6970 G-Skill Ripjaws DDR3 1600 
Hard DriveHard DriveOptical DriveCooling
SanDisk Extreme SDSSDX-240G-G25 Seagate Barracuda 7200.11 LITE-ON Black 12X BD-ROM Corsair H80 
OSMonitorKeyboardPower
Windows 7 Ultimate 64-Bit ASUS VS228H-P 21.5-Inch LED Monitor Logitech Wave Corsair TX750 V2 
CaseMouseMouse Pad
Corsair Carbide 500R Logitech MX-518 XTRAC PADS Ripper 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
AMD Phenom X6 1090T ASUS M5A97 EVO Sapphire HD 6970 G-Skill Ripjaws DDR3 1600 
Hard DriveHard DriveOptical DriveCooling
SanDisk Extreme SDSSDX-240G-G25 Seagate Barracuda 7200.11 LITE-ON Black 12X BD-ROM Corsair H80 
OSMonitorKeyboardPower
Windows 7 Ultimate 64-Bit ASUS VS228H-P 21.5-Inch LED Monitor Logitech Wave Corsair TX750 V2 
CaseMouseMouse Pad
Corsair Carbide 500R Logitech MX-518 XTRAC PADS Ripper 
  hide details  
Reply
post #9 of 40
Quote:
Originally Posted by NotUrAverageJoe View Post

I wouldn't suggest that. Corsair switched from Shin-Etsu to Dow-Corning - both are at worst equal to Arctic Silver 5. I use the stock TIM on my H80 and right now my idle temps are 27 core/ 35 degrees CPU sensor. I should note that Thubans read temps inaccurately at idle, but that's another story.
I should also note that I use Arctic 5 in my other build, so I'm not badmouthing it in any way. smile.gif
He's 100% spot on with that part. 9 times out of 10 thermal application is your cause for high temperatures.

I wasn't knocking the Corsair paste, I just remember them changing it from the Shin-Etsu stuff. I used the stock paste on my H80 for a good while, but the AS5 wasn't any better or worse really after I reapplied the block.

Both are good, it was just a simple recommendation that's easy to come by in times of needing to reseat your pump biggrin.gif
    
CPUMotherboardGraphicsGraphics
Intel Core i7-4790K ASRock Fatal1ty Z97X NVIDIA GeForce GTX 780 NVIDIA GeForce GTX 780 
RAMHard DriveHard DriveHard Drive
32GB ADATA XPG DDR3-2400 1x 256GB Plextor M6e M.2-PCIe 2x 256GB Samsung SSD 830 2x 2TB Western Digital Black 
OSMonitorKeyboardPower
Microsoft Windows 6.4.9841 CTP Qnix QX2710LED Logitech G710 EVGA Supernova 1000 G2 
CaseMouseAudio
Corsair Carbide 500R Logitech G3 Creative Zx, Sennheiser HD598 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel Core i7-4790K ASRock Fatal1ty Z97X NVIDIA GeForce GTX 780 NVIDIA GeForce GTX 780 
RAMHard DriveHard DriveHard Drive
32GB ADATA XPG DDR3-2400 1x 256GB Plextor M6e M.2-PCIe 2x 256GB Samsung SSD 830 2x 2TB Western Digital Black 
OSMonitorKeyboardPower
Microsoft Windows 6.4.9841 CTP Qnix QX2710LED Logitech G710 EVGA Supernova 1000 G2 
CaseMouseAudio
Corsair Carbide 500R Logitech G3 Creative Zx, Sennheiser HD598 
  hide details  
Reply
post #10 of 40
I actually think those temps might not be too far off. Not all Thubans have the craptastic core sensor. The BIOS and Asus probe are showing you the socket temps which are the temps at the CPU socket but not the temps of the CPU itself. These will always be higher than the core temps unless something is really wrong with your CPU. So if your BIOS, HWMonitor and Asus probe are all showing the socket temp at 64 at max load and 50 on core temps, that might be about right.

Mine show 60-61 on the socket and 50-54 on the core under Prime95 and I have a similar overclock to yours. You have a much better cooler so that would explain the 4 degree cooler reading youre getting on the core.

If youre not going to overclock the NB, try turning that voltage down. Cutting that made a HUGE difference in my temps.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD CPUs
Overclock.net › Forums › AMD › AMD CPUs › ★ ˜1090t Overclocking/Temperature Confusion ★