• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

RealTemp General Discussion

hey unclewebb, you should just link members to the information you contained in a PM you sent me ages ago. That sure changed my tune and take on the 45nm architecture in one sweet PM :)

Re: Thermal throttling

--------------------------------------------------------------------------------

I think most people don't realize how durable these chips are. 95C is not the thermal shut down point for these chips. That doesn't seem to happen until 125C! The words "holy sh!t" probably comes to mind.

125C is fully documented by Intel for the mobile chips but they don't list anything for the desktop processors so one user did his own testing and sent it to me. He got pretty damn close to that number but his computer would always shut down. Here's his maximum screen shot:

procthotbg0.png


RealTemp was still going strong at a core temperature of 123C while CoreTemp and Everest were badly confused.

Another user has been testing his E8400 lately and found that even when it was so hot that it was thermal throttling, it still continued to run Prime stable even at 99C.

capture25072008224232bs2.png



I think these CPUs can take a lot of thrashing without any long term problems. Even crazy overclockers are going to have to work hard to kill one of these chips. As long as you're running stable there is no need to even be concerned about core temperatures.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

That PM alone took everything I ever thought about Intel CPU's and threw it out the door. Mussels OCCT Linpack test often takes my proc to near 80* and I don't even flinch anymore. Thanks a ton for the insight to a whole new realm of overclocking.
 
Last edited:
What did they do to get those temps, turn off their CPU fans? First one looks like he took off the heatsink altogether (although, it is mighty underclocked).

And if they were to have been pushing 1.4-1.45 volts? How would that work out @ 90+C? What would the life expectancy be?

Too many unanswered questions for me to be pushing more and more volts and more and more heat through my CPU. I'd really like to get 3.8 stable but hitting 80C load core temps in Linpack with 1.36V (1.4 set in BIOS) and still unstable, dunno bout all that!
 
I believe the first image is a Q with no heatsink, and that is actually the point of thermal throttling, hence why the multi and speed are down.

The second image was by another user and I'm thinking it was without as well. But both show the limits of both 65 and 45nm procs, and the fact that even if you do get to the throttle point, the CPU is still stable. SO it is BIOS settings and OCCT limits that shut the PC down, not the CPU!!!
Of course unstable is still unstable, but it isnt due to temps...lol

Directly about voltage. I have run 1.6ish ( I dont reccomend this to everyone, Every CPU is different) volts through these for benching, no real ill effects, but I do swap parts often. So its all relative to your plans. I can respect keeping a lower OC for longevity, but I really have no such plans for my procs, they are usually 6 months to a year old when I replace them.
 
Last edited:
I think most people don't realize how durable these chips are. 95C is not the thermal shut down point for these chips. That doesn't seem to happen until 125C! The words "holy sh!t" probably comes to mind.
I beat you with another screenshot. :p

30dijis.jpg
 
I had to go back a few pages but here's my favorite test of what an Intel CPU can take.

I wanted to find out what would happen if while I was running a high stress program like Prime 95 Small FFTs that suddenly the fan died. The easiest way to simulate that is to pull the plug out of the motherboard and then go and watch TV.

What happened? Click on the following link to find out. :)

http://forums.techpowerup.com/showpost.php?p=1093655&postcount=97
 
Images like that are what keeps me trying:)

Loved the thermal status...HOT...ummm yeah I'd say.
 
The good news is that for 3 hours it just flat lined at 98C. When it got a tiny bit hotter, the CPU would automatically drop the multiplier down to 6.0 to start thermal throttling and to reduce the heat. When it cooled off half a degree, it would send the multiplier back up to 9.0 to resume running at full speed. This is happening hundreds of times a second. The Intel Core CPUs do an incredible job of looking after themselves. Just make sure TM2 is enabled in the bios and you'll never have to worry about a CPU fan failing. Many users might not even notice other than their computer might be a little quieter. I could probably game all day without a CPU fan and never hit the thermal throttle.
 
Does this look OK??

I just finished the calabration testing of my CPU, a q6600 G0. I have posted my results and am wondering about the LARGE difference between the TJMax and my acutal temps. Also, I have had issues with my MCP temp going above 78C, now from what I have read here, this is no issue, or is it?
Is that temp even taken into consideration for any of the alarms stepping etc? Should I even worry about it? Can you please take alook and see if there are any issues with my sensors. I could not get the image to show via the insert pic, so I have attached it

Thanks Much

Glenn :)
 

Attachments

  • Prime95 sensor resultsCut.jpg
    Prime95 sensor resultsCut.jpg
    108.9 KB · Views: 497
I just finished the calabration testing of my CPU, a q6600 G0. I have posted my results and am wondering about the LARGE difference between the TJMax and my acutal temps. Also, I have had issues with my MCP temp going above 78C, now from what I have read here, this is no issue, or is it?
Is that temp even taken into consideration for any of the alarms stepping etc? Should I even worry about it? Can you please take alook and see if there are any issues with my sensors. I could not get the image to show via the insert pic, so I have attached it

Thanks Much

Glenn :)

Load your cores with Orthos/Prime95 CPU stress/Small FFT or Linpack (Apps: LinX, IntelBurnTest, OCCT) and tell us the temps then. IMO the idle temps look too high, unless the ambient temp is like 82F.

Also, fill out your system specs; you can get there through the User CP. Look at the Control Panel on the right hand side.
 
The tests showen were with prima95, If you look at the top of the far right panel you will see the 98.7 percent load temps, all the way down to the idle. My main question is this, my room temp is 70F, these numbers are with the side panel of the computer open. Which app are you talking about? None of the apps have any CP on the right hand side. I built this system 5 months ago, at that time the CPU temps were being monitored by speedfan, in Vista Ult. I am now running Windows 7 Ult, the temps went up about 10C 2 weeks ago. I sent an email to the manufacture and they came back telling me this:
Dear Sir,

I'm afraid, that Alpine 7 won't be able to bring better temperatures, since the Q6600 has a TDP of 105 Watts, whilst Alpine 7 only brings 90 Watts of cooling performance.

Freezer 7 Pro or Freezer Xtreme should bring a much better performance for your applicatoin.

Best regards / Mit freundlichen Grüßen

After looking my cpu up, I found that the Max rated was 105, but according to INTEL with the stepping enabled the chip only needs 95 watts, since these butthead's did not include this information with thier spec sheet on the cooler, I am turning them in for fraud, and they can sink to the bottom of the earth before I will EVER use any of thier products again.


EVGA 132-CK-NF78-A1
Intel Quad Q6600 2.4
4 gig Adata PC 800
AC Pro 7 Air cooled
120mm intake fan
1- 80mm exhaust
Asus 9800GT w/512mg DDR3
WD SATA 320GB
Seagate 250 GB PATA 100
ASUS DRW1814-BLT SATA DVD-ROM
Mid Tower
850 Watt Cooler master PS

Glenn

P.S.
1.6 percent load 36, 36, 34, 35. the system has been sitting at these temps for over an hour
 
Last edited:
OK I just noticed that. Your temps are great then.
 
Thanks, at least now I can use this system without wondering if I am going to fry it. I am still a bit confused on how the calabration is supposed to work. I will read the post again and see if I can get the temps to read correctly.

Thanks again
Glenn:toast:
 
Hey unclewebb.

Does RealTemp write any registry intries? Like what folder it got installed to?
If it does, could you tell the reg path for it?
 
I hate programs that write anything to the registry. In my opinion, the registry is a weak link in Windows so I avoid it.

You can drag the RealTemp folder wherever you like. As long as RealTemp and the WinRing0.dll are in the same directory and you have the necessary permissions to execute files in that directory, RealTemp should run fine.

To get RealTemp to start up with Windows, you either have to drag a link to RealTemp into your Starup folder, add an item to your registry start up area or if you use Vista or Windows 7, you can use the Task Scheduler which gives you full control of items that start up on your computer. I like being in control of the endless junk on my computer. :)
 
I hate programs that write anything to the registry. In my opinion, the registry is a weak link in Windows so I avoid it.

You can drag the RealTemp folder wherever you like. As long as RealTemp and the WinRing0.dll are in the same directory and you have the necessary permissions to execute files in that directory, RealTemp should run fine.

To get RealTemp to start up with Windows, you either have to drag a link to RealTemp into your Starup folder, add an item to your registry start up area or if you use Vista or Windows 7, you can use the Task Scheduler which gives you full control of items that start up on your computer. I like being in control of the endless junk on my computer. :)

Dude, Registry has been the main point of Windows user data storage for YEARS. You still wish they wrote everything to INI files Linux style? IMO with the RAM and processing power in systems these days, the Registry has usurped and surpassed single-file based user data storage.
 
the advantage to keeping the settings in the folder with the exe, is that they transfer when you move the folder.

You can format, and still keep your settings/calibration etc.
 
the advantage to keeping the settings in the folder with the exe, is that they transfer when you move the folder.

You can format, and still keep your settings/calibration etc.

that's all gone down the shitter these days. you want your system backed up, either the app has an internal way to do it, you backup your entire system/registry, or you start fresh, and take pride and joy in knowing exactly how to set everything the way it was before you nuked it, like a real pro :D
 
I don't like programs with registry entries either. At least for a tiny app like RealTemp is perfect fine as it is now.
 
Dude, Registry has been the main point of Windows user data storage for YEARS.

Don't worry, I'm not that dumb. I realize that the registry is the center of windows data storage. That doesn't mean it's any good or doesn't have some serious issues. For a simple monitoring utility like RealTemp, it just doesn't make any sense to be storing information in the registry. A single INI file works fine. That way the entire program and all of its data is in one simple file that a user can drag to any location he likes and if he changes his mind a day or a week from now, he doesn't have to worry about uninstalling the program and reinstalling the program and wondering how much junk got left behind in the registry.

Any user can open an INI file and quickly edit an entry without having to worry about screwing up Windows in the process. I guess I learned programming with the Amiga OS that didn't have a registry so I never got in the habit of using a registry. Given that an old 16 MHz Amiga can do a full boot up in less than half the time that a 5 GHz Windows 7 machine can boot up, I'm just not that impressed by the bloated Windows OS and its registry.
 
i prefer the registry to be used as little as possible. its the main reason you need to format an old OS, is cause the registry gets bogged down and corurpted by bad programmers making bad programs with worse uninstallers.
 
I just came across this TomsHW article about TJ Maxes and I'm just wondering what you guys think of it.

According to this list, my RealTemp temps should be reading 14/12 idle. If it's correct I'd be insanely happy with my Thermaltake V1. :rockout:
 
That is one misleading document. Intel got a few of them right but when they started releasing data for the old 65nm chips, there was some major confusion. They came out with an updated version after myself and others complained about how absurd their numbers were but even the updated version wasn't much better.

What CPU model do you have?
 
E6750 G0

I've read that the TJ Max is 95 elsewhere
 
http://processorfinder.intel.com/details.aspx?sSpec=SLA9V

http://processorfinder.intel.com/details.aspx?sSpec=SLACR

Regardless of what Intel claims, core 0 and core 1 on my Q6600 are TJMax = 100C. Your E6750 and my Q6600 are both G0 Stepping, they have the same CPUID (06FBh) and the Thermal Specification is actually 1C higher for your CPU (72C vs 71C)

I'm pretty sure that TJMax for your CPU is 100C as well.

The best way to test if your choice for TJMax makes sense is to lower your core voltage in your bios as low as it can go and lock it to (6x333) ~ 2000 MHz. You can also enable EIST and C1E in the bios and it should drop down to these settings at idle. Check your voltage core with CPU-Z.

If you're not too far from 1.10 volts, your reported core temperature should be about 6C above your room temperature at idle near your open computer case depending on what heatsink and fan you're using. It usually becomes pretty obvious that 95C is too low.

The fact of life is that TJMax is not a fixed number. Intel says there is some variation from one part to the next but has never told users how much variation there is. Doing my test will help get you in the ball park. Post some screens if you do some testing.

I just updated my RealTemp / RivaTuner plugin tonight for anyone interested in that.
http://www.fileden.com/files/2008/3/3/1794507/RTCore.zip
 
Back
Top