![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Core temp or CPU temp?
- Thread starter trick0502
- Start date
Core temp is a better program hands down, simplier too.
version 0.95.4
Speedfan has the temp bug where readings are like 15C less than in reality. or something.. (Old version)
I suggest you try coretemp, cause those temps seem high.. How high is your overclock? vcore?
version 0.95.4
Speedfan has the temp bug where readings are like 15C less than in reality. or something.. (Old version)
I suggest you try coretemp, cause those temps seem high.. How high is your overclock? vcore?
I am having trouble quite comprehending what the question you are asking is. You might want to elaborate a bit more. You say your CPU temp is 62C, then ask if that is CPU temp or core temp. But it sounds like you just said it was CPU temp. So I got confused.
Are you having trouble reading the temps in Speedfan? Sometimes the CPU temp isn't labeled well, but the core temps should be easily identifiable. The CPU temp is usually a few degrees lower than the core temps (assuming the core temps are being read correctly by the program), so usually easy to figure out.
Another way I could of read this post is that someone told you the max temp of the Q6600 is 61 or 62 degrees and you were trying to sort out if they meant the CPU temp or core temp. Well they probably meant core temp since that is generally what we pay attention to. However 61 or 62 degrees is on the conservative side. But what some people will tolerate is a matter of preference. Some are okay with the core temps getting up to 65C, 70C, 75C. Most might want to avoid 80C or higher though.
Speedfan beta version 4.34 I believe is correctly reading the core temps on the G0 chips like the newer Q6600. My E6750, also a G0 chip, is now showing the correct temp in the beta version of Speedfan 4.34 I downloaded too.
Are you having trouble reading the temps in Speedfan? Sometimes the CPU temp isn't labeled well, but the core temps should be easily identifiable. The CPU temp is usually a few degrees lower than the core temps (assuming the core temps are being read correctly by the program), so usually easy to figure out.
Another way I could of read this post is that someone told you the max temp of the Q6600 is 61 or 62 degrees and you were trying to sort out if they meant the CPU temp or core temp. Well they probably meant core temp since that is generally what we pay attention to. However 61 or 62 degrees is on the conservative side. But what some people will tolerate is a matter of preference. Some are okay with the core temps getting up to 65C, 70C, 75C. Most might want to avoid 80C or higher though.
Speedfan beta version 4.34 I believe is correctly reading the core temps on the G0 chips like the newer Q6600. My E6750, also a G0 chip, is now showing the correct temp in the beta version of Speedfan 4.34 I downloaded too.
speedfan 4.34 and coretem report the same core temps (4.34 reports the correct temps, 4.33 doesnt). what im wondering when you say a cpu has a max temp, i that temp the temp of the hottest core, or of the cpu overall?
see haw the cpu is 54 and the hottest core is 67? i know the max temp for my b3 q6600 is 61, is intel talking about the core temp or the cpu temp?
see haw the cpu is 54 and the hottest core is 67? i know the max temp for my b3 q6600 is 61, is intel talking about the core temp or the cpu temp?
WoW!
What is your overclock and how much Vcore are you giving it. Those tempartures are way to high. (Dependant on the information you have not given us)
I want to know the Vcore you set through the CMOS, not some software read out.
I would seriously just use Coretemp 0.95.4
Be more concerned about the Core temps..
What is your overclock and how much Vcore are you giving it. Those tempartures are way to high. (Dependant on the information you have not given us)
I want to know the Vcore you set through the CMOS, not some software read out.
I would seriously just use Coretemp 0.95.4
Be more concerned about the Core temps..
Those temparture are absolutely horrible for the clock and Vcore of your CPU.
What thermal paste did you use? Are you sure the heatsink is properly secured? What is the room temparture? other case cooling? Amount of fans?
I would most definitly check the condition of the heatsink and thermal paste at this point.
I just hit 61C full load, with my Q6600 @ 3.7 Ghz...
What thermal paste did you use? Are you sure the heatsink is properly secured? What is the room temparture? other case cooling? Amount of fans?
I would most definitly check the condition of the heatsink and thermal paste at this point.
I just hit 61C full load, with my Q6600 @ 3.7 Ghz...
Its a G0, But your tempartures are still too high for the current clock and Vcore, way way too high..
I reccomend lowering room temparture, new heatsink and perhaps even MX-2 thermal paste.
Heatsinks such as but not limited to.
Scythe Mine
Scythe Ninja
Ultra 120
TRUE
You may even want to consider lapping your processor.
I reccomend lowering room temparture, new heatsink and perhaps even MX-2 thermal paste.
Heatsinks such as but not limited to.
Scythe Mine
Scythe Ninja
Ultra 120
TRUE
You may even want to consider lapping your processor.
Which was already answered.
The core temps. I would ignore what Speedfan says about the CPU temparture, just pay attention to the four Core temp read outs.
And please, just give core temp 0.95.4 a try....
Goto a automative or paint store for higher sand paper grits.
And you will still want to lap the IHS more than the heatsink base. (You can still do the heatsink base, but it your lapping anything the IHS is a must)
You need a brand new razor blade to test for flatness, and even if you say so, the unlapped IHS is never perfectly flat or without defects...
I get very annoyed when people do not answer questions....
What is your room temparture?
How many case fans do you have?
The core temps. I would ignore what Speedfan says about the CPU temparture, just pay attention to the four Core temp read outs.
And please, just give core temp 0.95.4 a try....
Goto a automative or paint store for higher sand paper grits.
And you will still want to lap the IHS more than the heatsink base. (You can still do the heatsink base, but it your lapping anything the IHS is a must)
You need a brand new razor blade to test for flatness, and even if you say so, the unlapped IHS is never perfectly flat or without defects...
I get very annoyed when people do not answer questions....
What is your room temparture?
How many case fans do you have?
my room temp is cold, i dont know the exact temp, but my system temp is 32 if that is any help. i have an antec 900, 3 120mm, and a 200mm or 250mm fan on the top, so case cooling is great. to be honest its only about 2-4c cooler at stock speed, the cpu runs hot.
zero_vertical
Gawd
- Joined
- Jun 6, 2006
- Messages
- 1,001
the type of thermal paste doesn't matter. Studies have proven that type of thermal paste should have no effect on temps. Spending major $ on thermal goo is a sham. Just use the stuff you get with your processor.
His readout is borked, nothing else. These software monitoring programs are flaky.
Intel says ignore the core temps and just look at the "CPU Temp". Their temp limits are all limits on that temperature (Tcase). The core temps are only used in mobile processors where the diode on the IHS that measures Tcase is not present, because there is no IHS at all.
More detailed info (see chapter 5):
http://download.intel.com/design/processor/datashts/31559205.pdf
More detailed info (see chapter 5):
http://download.intel.com/design/processor/datashts/31559205.pdf
Then.. that means my CPU is only 28*c-31*c idle .. OCing from 2.4 - 3.0..
That can't be right, can it?
If you're just using stock voltage and the room is cool, those are normal numbers. That temperature is supposed to be calibrated to match the temp at the top side of the heat shield, at the center, so during idle its perfectly ok for that temp to be only a few degrees above your case temp.
I'd like to see those links. Every study I've seen has indicated that the type of thermal paste does have an effect on temps.
And to the OP: Will you stop talking about your CPU temp? Like what most people have tried to tell you, it's the 4 core temps that you should be concerned about and they are too high. And as previously mentioned, perhaps a contact problem between your CPU heatsink and CPU. However, I suggest simply removing your heatsink, reapplying TIM, and resetting it.
Please don't take offense to this but, I find that to be a load of crap, sure the differences may be minor, but there is still a difference between thermal pastes.
http://translate.google.com/transla...fe=off&ie=UTF-8&oe=UTF-8&prev=/language_tools
Battle of the thermal Pastes, Non original Post
http://www.tech-forums.net/pc/f76/battle-thermal-compunds-144331/
The point is this.
"
|crappy thermal interface|
Tital Nano Blue 59
Panasonic 58.5
KPT-8 (reference) 56.6
|decent thermal interface|
Data Cooler 57
Titan TTG S-104 , S-103 56.7
Pasta Siliconowa 56.6
Zalman CSL 850 56.5
Noctua 56.5
Stars Silver 56.5
Stars 700/Aero 700 56
GeIL GL-TCP1b 55.6
Thermopox 55.6
КПТ-8(BeO) 55.5
Sil more 55.5
Shin-Etsu(white) 55.5
W.P. 55.5
STARS (white, soft pack) 55.5
AKT-842 55
Fanner 420 55
Koolance 55
|good thermal interface|
Arctic Alumina 55.5
Arctic Silver-3 54.6
AOS 54.5
DC-340 54.5
Asetek 54.5
Arctic Silver-5 53.5
Arctic Ceramique 53.5
|great thermal interface|
Apus TMG 301 52.5
Gigabyte 52.5
Titan Nano Grease TTG-G30010 52.5
GFC-M1 D90T8-010 52.5
Shin-Etsu MicroSi G-751 52
|outstanding thermal interface|
Arctic Cooling MX-1 51.5
Arctic Cooling MX-2
Shin-Etsu MicroSi MPU-3.7 50.5
Coollaboratory Liquid Pro 50
"
If there was no difference between "Panasonic 58.5" and "ARCTIC COOLING MX-2" I would not have "wasted" my money...
well according to intel its the cpu temp not the core temp
I don't trust intel for much....
Other than for their nice overclocking processors that is..
Then why ask the question if you already know the answer?
Anyway, I'm curious and don't have the patience to research the Intel documents like you obviously did. Quote me the section where Intel says this says to use the cpu temp rather than the individual core temps.
(Randomly looking through that Intel document I see (for dual core): "An external signal, PROCHOT# (processor hot), is asserted when the processor core temperature has reached its maximum operating temperature. PROCHOT# (Processor Hot) will go active when the processor temperature monitoring sensor detects that one or both cores has reached its maximum safe operating temperature....The PROCHOT# signal is bi-directional in that it can either signal when the processor (either core) has reached its maximum operating temperature"... Sounds like it checks each core...)
TheRapture
Supreme [H]ardness
- Joined
- May 31, 2000
- Messages
- 6,885
I can guarantee you, given PROPER application and clamp pressure on the heasink, the difference between WHITE oem type paste, and AS3/5/etc., is no more than 2c-3c. You can read all the "tests" you want, but many are biased (advertiser bias) and many are poorly controlled. I and many others, have tried both, and after years in the hardware scene...there is simply little difference.
AS5 and the likes do last longer without drying out like the white stuff....that's about the only real advantage. 2c-3c is not going to make an unstable overclock, become stable, or vice versa.
AS5 and the likes do last longer without drying out like the white stuff....that's about the only real advantage. 2c-3c is not going to make an unstable overclock, become stable, or vice versa.
your saying the zalman 9500 is not a good heatsink/fan? i mean i know there are better but come on i dont think he needs a new heatsink for that kind of overclock
2c-3c is a difference, especially when stated in the context of a forum where people polish their CPU's and heatsinks, and obsess about the heatsink used, simply to get slightly lower temps. Blueeagle has indicated the difference can be up to 9c depending on the TIM used, and backed it up with a study. Can you back up your claim of "2c to 3c"?
Here's another study, from MADSDHRIMPS:
To the OP: If you don't know the answer, then...ah nevermind. Entirely up to you, but may I suggest you use this utility when comparing temp. It's a standard utility that appears to have far greater acceptance amongst overclockers for comparing temps than speedfan. It's called "coretemp" and can be found here : http://www.alcpu.com/CoreTemp/
In your opnion...I find the Zalman 9700 does the job like a charm too...
TheRapture
Supreme [H]ardness
- Joined
- May 31, 2000
- Messages
- 6,885
echn111....ease off the ego bud. It's just thermal paste, and if properly applied....like I said.....2-3c is about the MOST AS5 and white oem paste will give you. If there is a 9c difference, something was done WRONG, period. Maybe you need to learn how to aply paste properly?
BillParrish
Supreme [H]ardness
- Joined
- Aug 25, 2006
- Messages
- 7,519
Woo HOO !!! one of "these" posts ! Big fun.
To the OP,
The CPU temp reported by speedfan is from a sensor located somewhere on the motherboard (probably in the little "well" formed by the square hole in the middle of the socket) probably under the CPU. This is a thermistor and is not very accurate and responds slowly.
The coretemps are readings from the internal temp sensor system built inside the CPU circuitry and are much more accurate and very sensitive and fast.
To test, watching speedfan, kick on Orthos or some other load software when the machine has been a idle for a while. Coretemps will instantly shoot up, the CPU temp will rise much slower. Let it run a bit, then turn off Orthos, coretemps instantly drop, but CPU temp drops slower.
Coretemp is what you want to watch. Your coretemps are fairly high for your CPU and setup, no clue why but something seems amiss. However you have absolutely nothing to worry about until coretemps reach above 80C and actually do not even worry then, It is almost impossible to damage a C2D with high temps, it will just shutdown. If you would like a full explanation, only mildly technical, but a lot to read, look at my post here: http://hardforum.com/showpost.php?p=1031827510&postcount=31
Footnote: it is with deep saddness I report the world famous, absolutely classic, "study", that was once posted on rabid reviews, that without a shadow of a doubt proved that creamy peanut butter was just as good as AS3 or anything else out there at the time, is no longer on the web. /sniffle.
To the OP,
The CPU temp reported by speedfan is from a sensor located somewhere on the motherboard (probably in the little "well" formed by the square hole in the middle of the socket) probably under the CPU. This is a thermistor and is not very accurate and responds slowly.
The coretemps are readings from the internal temp sensor system built inside the CPU circuitry and are much more accurate and very sensitive and fast.
To test, watching speedfan, kick on Orthos or some other load software when the machine has been a idle for a while. Coretemps will instantly shoot up, the CPU temp will rise much slower. Let it run a bit, then turn off Orthos, coretemps instantly drop, but CPU temp drops slower.
Coretemp is what you want to watch. Your coretemps are fairly high for your CPU and setup, no clue why but something seems amiss. However you have absolutely nothing to worry about until coretemps reach above 80C and actually do not even worry then, It is almost impossible to damage a C2D with high temps, it will just shutdown. If you would like a full explanation, only mildly technical, but a lot to read, look at my post here: http://hardforum.com/showpost.php?p=1031827510&postcount=31
Footnote: it is with deep saddness I report the world famous, absolutely classic, "study", that was once posted on rabid reviews, that without a shadow of a doubt proved that creamy peanut butter was just as good as AS3 or anything else out there at the time, is no longer on the web. /sniffle.
Careful about bringing emotion into what should be a logical argument. And if you think I did so, re read the posts. My point was that, unlike you, the other poster backed up his assertion. I agreed with the other poster, and then further backed it up with another link and a pic.
Perhaps my view is wrong, and that's fine, just say so and back it up. Unfortunately, you on the other hand, have continued to back up your assertion with "nothing". To make matters worse, you then try to distract from this by making a blatant emotional dig. So rather than acting like a fool, if you had given some evidence that what you're saying is right, I'd respect that. As it is, you've lost whatever credibility you had.
Bill,
There are three temperature sensors on C2Ds and five sensors on C2Qs. The on-die DTS which gives each core temp and the "center of the processor" sensor that approximates the "Tcase."
The reason the third/fifth sensor lags the core sensors is because the IHS does not heat/cool as quickly as the cores.
Its not an opinion. Its fact, there ARE better heatsinks than the Zalman 9700, and there ARE worse heatsinks than the zalman 9700.
The Zalman 9700 is neither "that" good OR "that" bad. Thus what I originally said.
If one made the statement that the 9700 is the best heatsink ever they would be WRONG.
If one made the statement that the 9700 is the worst heatsink ever they would also be WRONG.
And I know you didn't say it was the best or the worst, but I am just pointing out that what I said was not opinion. I Know for CERTAIN that "its not bad, but its not that great either.."
And it is in fact YOUR OPINION, that the zalman 9700 "does the job like a charm[.]" Thats fine, I have no arguement with that. All I said is that there are BETTER, and once again, that is a fact, NOT an opinion..
I repeated myself alot to get a point across. I hope I succeeded..
Oh, and doesn't the OP have a 9500 and NOT a 9700??