Discussion in 'RealTemp' started by Sxx, Apr 26, 2010.
Yeah you should stick around. You'll learn loads.
temps r great push that thing to 3.4
VERY RESPECTABLE TEMPS!!! Now from what I see you have some overclocking headroom if you desire!
idle temps at 3.6ghzs 400x9 cpu ratio
full load prime 95 then i ran 3dmark vantage here are the full load temps
I must say my pc does feel very fast indeed now very fast
I feel all excited and i cant describe the other feeling i won`t go any higher i think it fast enough now
i turned up all my antec 900 fans to full now
i love this website and you guys for giving me the knowledge and courage too
u sir got a good chip and great job OCing
Now try 3.8 hehe
thank you all
i have ordered another fan to put on the back of the heatsink i will cable-tie it or use a band to hold it in place
I`m scared to OC it any higher my asus p5q deluxe came with a mini asus fan i have attached it to the heat pipes it said for use with a passive cooler or water cooler.
Here's a fully loaded Q6600 running hot as hell.
It hasn't even started thermal throttling yet because Intel considers this temperature to be OK. If your CPU ever runs too hot the OK word in RealTemp will change to LOG. That will mean that at least one thermal throttling episode has been logged. As long as you keep it under that temperature you are fine and it will continue to run at full speed.
When overclocking, you will run into stability issues long before you ever run into temperature issues. As long as your computer is stable and not thermal throttling, you don't have to give the core temperatures a second thought.
Edit: And this is a sign that you've crossed that line and gone too far.
The word HOT means that thermal throttling is in progress and your CPU is not running at full speed. That's no good.
I like testing these features so you guys don't have to worry.
Wow that is hot ^.^
thank you mr unclewebb for the testing
Most i have hit is 62c core temp across all 4 cores... May i ask does your q6600 still work though ????
thank you for taking the time to post that and to all the other people that have replied and gave advice/knowledge thank you benjamin
so TJMAX 100C is really alot higher before it thermal cuts down ???
Unclewebb!! woot, yea Sxx that's the guy to listen to forsure. btw, congrats on your OC, surpassed my max of 3.4ghz on this mobo and you've just started oc'ing. See nothing to her
It did still work long after those screen shots were taken. It didn't seem to make any difference. I sent that Q6600 to a friend in Australia 6 months or so ago and I haven't heard any complaints so I assume that it's still working fine. Whenever something stops working, you usually hear about it. I think he went and abused it with more voltage than I ever gave it. I think it's more important to keep the voltage reasonable and within the Intel specs. Core temperature doesn't seem to be a big issue.
The thermal throttling point is 100C for a Q6600 - G0 stepping where the multiplier will start to rapidly cycle back and forth between 9 and 6 which slows the CPU down which helps to cool it. The thermal shutdown temperature isn't until 125C to 130C so you can see that Intel has a lot of confidence in their CPUs.
thanks, didn't know that. now ima run my CPU at 3.8 with idle temps of 64 and load 85 celcius
My thoughts exactly a ump
Unclewebb thank you for clearing that up sir thank you they sure do take some heat them chips i dont care about temps anymore like you said if it gets to hot i will notice it shutting down if it doesn`t do that then dont worry
race you to 4ghzs i will try it when i get in from school tonight 4ghzs and beyond lol
why dont you send ME CPU's
lol i was jk'ing SXX. 3.4ghz is as high as i can go on this board. It's only CPU voltages are 5,10,15% percent of stock voltage increases, and i already require 15% increase to get 3.4ghz stable. Temps are fine...just won't go any faster without more volts. when loaded i'm only at 1.35v so if i ever get a Gigabtye UD3R or something then i'll get it faster with more juice . mite start shopping for one now actually lol
Intel CPUs are rated to take a lot of heat and still run reliably. If you are overclocking, you will need to run them a lot cooler to maximize your overclock while maintaining stability. The above Q6600 could also run reliably at 3.2 GHz but I couldn't run that speed at those temperatures. As the speed goes up, you will be forced to keep the heat down or you will lose long term stability. All CPUs and especially the Core 2 based CPUs follow this pattern.
The original 45nm Core i7 CPUs seem to be able to overclock reliably even up into the 90C range. You won't be able to get a maximum overclock out of a Core 2 CPU while running at that temperature.
That's why you don't have to be too concerned about an exact temperature number when overclocking. Just run them as cool as possible and as long as your CPU is stable then you won't be anywhere near the Intel thermal specifications. Overclocking will force you to run your CPU cooler to get the maximum out of it. An overclocked Core 2 will either error out in Prime95 when it gets too hot or it will simply reboot.
I don't think I got a very good deal for that Quad. I lost two cores because I traded it for an E6600 and then I lost some MHz and some cache by trading the E6600 for an E4300. Collecting data to make RealTemp better isn't cheap.
But Intel have stated that the TJmax for the 65nm Q6600 is 90°C not 100°C.
So I'd enter 90°C into Coretemp/Realtemp as the TJMax.
Or just use Speedfan as that uses 90°C as per Intel's spec.
You need to read the fine print of Intel's big news release at their IDF Conference. Those numbers are not TJMax. They are TJ target and are completely meaningless. They admit in this document that actual TJMax may be different than the numbers listed. In many CPUs I tested, these numbers weren't even close to actual TJMax.
The page of the document you posted was so ridiculous that when programmers and users and myself complained, Intel released a second document to "fix" some of their errors. The programmer of Core Temp also ignored this document.
Use whatever TJMax makes you happy but you're better off picking a number out of a hat compared to using any numbers from the document that you posted.
The original poster has a Q6600 - G0 stepping which I tested. TJMax wasn't anywhere close to 90C. Actual TJMax for Core 2 and Core 3 was closer to 105C or 106C
I have plenty of stock clocked Q6600's that run in the high 80's low 90s on stock intel coolers on intel boards with fast good fans...
Your temps are solid
Thank you, i left mine at what ever the realtemp set it to i trust them more
Here's the formula that all software uses:
Reported Temperature = TJMax - Digital Sensor Reading
You don't have to be a math major to see the motivation for Intel to release some low ball TJ Target numbers. If TJ Target is 10C less than the actual TJMax and users adjust TJMax lower in RealTemp, they'll create what appears to be a cool running CPU. The result, less complaints about hot running CPUs.
It was a great plan but they got a little carried away when they released the above document. The first line that shows an E6000 with a TJ target of 70C is nice but it is 20C less than the actual TJMax for the E6400 and E6600 B2 stepping CPUs that I tested. When they updated that document, after asking me what my number was, they decided to split the difference and they used a TJ Target of 80C which is still 10C less than the actual TJMax.
I had already lost faith in getting an honest answer by that time. When I noticed that they were interchanging TJ Target and TJMax; I decided to take their flashy PR presentations with a grain of salt. They've never had the guts to publicly release any meaningful engineering data about the temperature sensors they use.
Ahhhhh ok matey
How do I go about calibrating my Q6600 then unclewebb? Something you can help me with good sir?
If it helps here my CPU sensor test and cool down test using P95 blend (is this correct, should I have use small FFT's?)
Separate names with a comma.