Who screwed me over..Intel or Tuniq??

For grins, here's a screenie of my rig, E6750 clocked at 3.2ghz....I use a Thermaltake Big Typhoon. Case is a Tsunami (thermaltake) and has medium airflow, nothing special. My temps as reported by coretemp, I use the "delta remaining to tmax" setting, and I am running TAT which loads the cores as hard as anything out there. I am STILL 30c+ away from hitting the max temp...that's a long ways.And DO NOT trust the temps that TAT reports, that software was designed for Mobile chips and DOES NOT report C2D temps correctly.
Have you guys tried using coretemp to show the delta? It's accurate.

7xeb5mf.jpg
 
Coretemp .96 is the latest........94.x is pretty old......why don't you try the new one?

Many users here show off lower temps with the older 0.95.4 ver. I am using the 0.96 ver.. thing is 0.96 reports higher temps for many Core 2's esp. the newer steppings (M0 for one).
Anyways, just for kicks my overclock is given below ..it is running 2 instances of Prime95 that generate about 5-6'C more heat that at 'normal' or a real world full load scenario.

oc.jpg
 
I know this is a dumb question, but did you apply the AS5 per the instructions on their webpage? That made a difference of 10 degrees when i applied too much AS5.
 
Thanks everyone for at least not leaving me to think I'm the only sucker around with lousy temps :D

Here is my story in short:
I have GA-P35-DS3R and E4400, stock clocks and stock Intel sink gave me a bit high results with TAT and Core Temp, and when I'd OC to 3.2 GHz it would be unbearable to watch at temps (in TAT at 100% load I'd go over 80C) even though they've relly gone up for like 5C vs stock 2GHz.
So I've tried changing paste with Arctic Ceramique, reapplying several times, and finaly I've even bought Ninja Rev 2. But even so, with CPU @ 3200 (but idle), I have Core Temp and TAT show one set of numbers (~45C), all others say ~30 including Gigabyte's own EasyTune updated today. I've also got big 250mm side panel fan pushing some serious air, it's with heating turned off in my flat, and outside temps under zero C. I get chils on my feet, just from the air going through the case and out on the other side! I've touched Ninja at load (80C) all over and I can barely feel some heat no matter where I hold on to it (fins, base, heatpipes on both bottom or top).. And my hands are real cold, so I should feel some real heat on 80C, right?

Btw, CPU is stable (OCCT and others, didn't run Prime) on minimum voltage increase (like 1.3825 is enough, even though it sometimes works for hours even on 1.3625V at OCCT load as well!). It just kills me to see 80-81C in TAT/Core Temp (~65C in other tools). "Tjunction max remaining" is at some 19C under full TAT load (maybe would go to 18 or 17 if I'd let it run for longer time) and 52-55C idle.

So this is fine than or not? Tomorrow I'll go and reapply Arctic Ceramique by instructions on AC site, and will than load CPU for 30 mins with both TAT and than Prime, with heating turned up and all, looking for worst case scenario.. will see what happens.

But as long as delta T is let's say 15-20C "to the max" it should be fine right? Right? :D
 
Core temp seems to display lower temps than what my bios's hardware monitor says. I'm sure your TRUE was applied correctly. People tend to see what they want to see (I'd rather have 50 degrees load on coretemp vs 60 degrees on my motherboard monitor).
 
Dunno ... never really gave it that much thought. Have always applied a thin layer covering the entire CPU .. then placed the HSF and slid it around a bit to spread it evenly.

Doesnt hurt to try .. will give it a shot and post.
 
FYI I have an E4500 running at stock with the stock Intel cooler. Idle temps are 45c. Full load is about 68c per the latest version of CoreTemp, so we are about the same temps. Only difference is I'm using the stock cooler. The fan on the stock cooler is running about 1500-1700rpm depending on if I go by what Everest is displaying or the bios. I have been trying to get this fan running at a higher speed but my bios has no fan control options what so ever. Even with quiet fan enabled or disabled (thats the only fan setting I have) I get the same fan speed no matter what my temps are.

It looks to me these temps are about normal even though they seem a bit warm.
 
FYI I have an E4500 running at stock with the stock Intel cooler. Idle temps are 45c. Full load is about 68c per the latest version of CoreTemp, so we are about the same temps.

Yours is the first C2D i can actually compare temps with. What is really interesting is the high temps though ... Core Temp does seem to show higher than usual temps for the M0 stepping E4500.

Fingers crossed and hoping that its Core Temp and not the CPU at fault, since i have exhausted all possible options to lower my temps.
 
Dunno ... never really gave it that much thought. Have always applied a thin layer covering the entire CPU .. then placed the HSF and slid it around a bit to spread it evenly.

Doesnt hurt to try .. will give it a shot and post.

According to the Arctic Silver website, you're just supposed to put a thin line down the middle (in between the 2 cores) and place the HSF on. Spreading it around doesn't sound smart to me, as you're can't be sure it's evenly spready. 45° idle/81°C load sounds kinda high. What are your ambient temps? Do you have any case fans?
 
For grins, here's a screenie of my rig, E6750 clocked at 3.2ghz....I use a Thermaltake Big Typhoon. Case is a Tsunami (thermaltake) and has medium airflow, nothing special. My temps as reported by coretemp, I use the "delta remaining to tmax" setting, and I am running TAT which loads the cores as hard as anything out there. I am STILL 30c+ away from hitting the max temp...that's a long ways.And DO NOT trust the temps that TAT reports, that software was designed for Mobile chips and DOES NOT report C2D temps correctly.
Have you guys tried using coretemp to show the delta? It's accurate.

Those temps are horrendous for that cooler, that's peaking at almost 70.

6+ hours of Orthos using stock cooler (CPU @ 3.4Ghz - also E6750) and I peak at 61.
 
According to the Arctic Silver website, you're just supposed to put a thin line down the middle (in between the 2 cores) and place the HSF on. Spreading it around doesn't sound smart to me, as you're can't be sure it's evenly spready. 45° idle/81°C load sounds kinda high. What are your ambient temps? Do you have any case fans?

I just tried this method too with absolutely no change in temps except for those in my cranium.. :mad::mad:

It makes absolutely no sense to me..hell i even did a great job of lapping both the Tower and the IHS.. I'm beginning to think this was all a total waste ..
 
"Tjunction max remaining" is at some 19C under full TAT load (maybe would go to 18 or 17 if I'd let it run for longer time) and 52-55C idle.

19c-20c under tjunction max is fine, decent even for a highly overclocked cpu, especially if you are in a warmer climate and/or heated room or low case airflow.

It's stable, you are sure you have good interface to the heatsink, and it runs good, screw it. You are within safe limits. :p
 
I just tried this method too with absolutely no change in temps except for those in my cranium.. :mad::mad:

It makes absolutely no sense to me..hell i even did a great job of lapping both the Tower and the IHS.. I'm beginning to think this was all a total waste ..

As far as any TIM goes, I have always used a straight razor edge to thinly spread paste on the entire heat spreader. Done right, I guarantee you it works as well or better than any other method. No over flow, no drip.
 
Those temps are horrendous for that cooler, that's peaking at almost 70.
6+ hours of Orthos using stock cooler (CPU @ 3.4Ghz - also E6750) and I peak at 61.

There is a HUGE difference between TAT and Orthos/Prime/OCCT tools.. I've just took a look, and TAT after ~20min 100% load gives me 83C (delta Tmax is 17 according to Core Temp) and after running OCCT 30min (don't have Orthos, looking for links right now!) test highest temp recorded was 28 delta Tmax. So his "almost 70" with TAT and your "peak 61" with Orthos means he actualy has BETTER temps than you, or there abouts.

19c-20c under tjunction max is fine, decent even for a highly overclocked cpu, especially if you are in a warmer climate and/or heated room or low case airflow.
It's stable, you are sure you have good interface to the heatsink, and it runs good, screw it. You are within safe limits. :p

Well I've remounted heatsink again, even took some photos of TIM when I took it off, and all seems fine. I had to wiggle it a bit to take it off, just to mention - since some ppl said if it goes off easily contact is probably bad. Also, I have mostly same temps after doing cleanup and applying AC Cermaique as said on AC site. So it seems that contact is good, and at worst under TAT I had delta Tmax 17C just for a moment there (with heating turned on in the room, though it's not summer's 40C at the moment :D ) while OCCT shows above mentioned delta of 28C. So even if it gets up to 35C in my apartment in summer, under real-world loads, I think I'll be fine.

But thanks for reassuring me a bit :D

Btw, I did check if IHS is flat or not. And it's NOT. Razor test shows it's concave, and I think that's part of the reason why Intel stock HSF has more or less same temps, as it's base is smaller and it gets closer contact with IHS in the center (it is closer to the "bottom" of the IHS "pit" I'd say :D ). But as you say, if 19-20C delta Tmax is fine, I won't bother voiding 3-year warranty of a new CPU :p

Only thing now is - why did I spend money on Ninja? :confused: Luckily, only paid it about 50$ (cheap around here, usualy it's over 70$! :p )
 
[email protected] with a Tuniq 120 here. 1.41vcore. Idles at around 23-25c. Under full load (orthos) hits about 47c tops.

Wow! Nice temps! Btw, could you measure your delta Tmax temperatures using Core Temp 0.96 - a) idle and b) after some 10 minutes being 100% loaded with TAT?

I think you've got all download links here in the thread if you don't have them..
 
19c-20c under tjunction max is fine, decent even for a highly overclocked cpu, especially if you are in a warmer climate and/or heated room or low case airflow.

It's stable, you are sure you have good interface to the heatsink, and it runs good, screw it. You are within safe limits. :p

Well, i'm done guessing what the higher temps are due to .. I'll live with it till i dump the E4500 for a quad. Any idea what is the TCase max for my M0 stepping E4500??

But as you say, if 19-20C delta Tmax is fine, I won't bother voiding 3-year warranty of a new CPU :p

The surface of my IHS is all copper now with no nickel pit. So hell yeah..i voided my warranty within a weeks time!
 
Wow! Nice temps! Btw, could you measure your delta Tmax temperatures using Core Temp 0.96 - a) idle and b) after some 10 minutes being 100% loaded with TAT?

I think you've got all download links here in the thread if you don't have them..

OK here's 2, I'm still using Orthos as my weapon of choice however ;)

orthoscoretemp096test.jpg


orthosloadcore096testtjunction.jpg


Oh and one at idle

idletempcore096.jpg
 
OK here's 2, I'm still using Orthos as my weapon of choice however ;)

Thanks! :) Ok, I'll let you slip on this ;) Btw, could you maybe point a link for some newer Orthos download?

But anyway, these are great :eek: Do you have lapped CPU perhaps? I mean, these are way way better temps than anyone posted yet, so that's only thing I can think of :D Or you keep your room at zero temps :D ;)
 
Thanks! :) Ok, I'll let you slip on this ;) Btw, could you maybe point a link for some newer Orthos download?

But anyway, these are great :eek: Do you have lapped CPU perhaps? I mean, these are way way better temps than anyone posted yet, so that's only thing I can think of :D Or you keep your room at zero temps :D ;)

Latest Orthos I was using was this. No the heatsink isn't lapped, just a thin layer of AS5 on the CPUI and the Tuniq tightened evenly. Good airflow through the case. My room isn't hot by any means, but is acceptable room temperature. I know how efficient the Tuniq and airflow is because when I stop Orthos, the CPU temp drops from mid 40s to mid 20s in a few seconds. That's a good measure of efficiently getting rid of heat :).
 
Those temps are horrendous for that cooler, that's peaking at almost 70.

6+ hours of Orthos using stock cooler (CPU @ 3.4Ghz - also E6750) and I peak at 61.

No, this cpu is an 85c tjunction MAX spec, meaning that since I have 31c REMAINING until I hit that temp, I am around ~56c under load. At 85c the cpu would shut down....at ~80c the cpu would start throttling, so even in worst case my load temps are still 26c DELTA before it would start throttling.

You are assuming that the E6750 has a max temp of 100c, it does not.

My temps are LOWER than yours.....which they should be since I am using a good cooler as compared to the stocker.
 
[email protected] with a Tuniq 120 here. 1.41vcore. Idles at around 23-25c. Under full load (orthos) hits about 47c tops.

dude, all mobo software temp reporting is known to be...pretty bad. I seriously doubt those temps unless you are water cooling or have extremely low ambient temps. without knowing ambient temps, cpu load temps are meaningless. Stop using the Abit software and use something like coretemp.....

edit: I see you did just that, and man, your temps look really low for air cooling...we need to know ambient temps and case airflow.... :)
 
dude, all mobo software temp reporting is known to be...pretty bad. I seriously doubt those temps unless you are water cooling or have extremely low ambient temps. without knowing ambient temps, cpu load temps are meaningless. Stop using the Abit software and use something like coretemp.....

OK. I'll refer you to this then ;)
 
Latest Orthos I was using was this. No the heatsink isn't lapped, just a thin layer of AS5 on the CPUI and the Tuniq tightened evenly. Good airflow through the case. My room isn't hot by any means, but is acceptable room temperature. I know how efficient the Tuniq and airflow is because when I stop Orthos, the CPU temp drops from mid 40s to mid 20s in a few seconds. That's a good measure of efficiently getting rid of heat :).

Thanks for the link, though I just stumbled on it myself :D
Btw, I've also found Prime 25.6 on some russian page (http://www.computerraru.ru/soft) and after some 10 minutes now, it's around 21/22C delta Tmax, even touched 20C for a moment. Not far from TAT's 17-18C, nice going!
Btw, it supports dual/quad cpus, so happy burning :D I'll leave it over the night tonight..
 
You are assuming that the E6750 has a max temp of 100c, it does not.

Pretty sure it is considering .95.4 and .96 both show tjunction as 100C for that chip, as well as everything I've read online. Can't find a spec sheet on Intel's site though mind you.

Been running TAT - peak temp is "28 to tjunction remaining", quite a bit hotter than Orthos for sure but even using the same settings you are in CoreTemp for temp recording your temps still look REALLY high to me since both yours and mine report Tjunction as 100C.
 
Pretty sure it is considering .95.4 and .96 both show tjunction as 100C for that chip, as well as everything I've read online. Can't find a spec sheet on Intel's site though mind you.

Check out extremesystems forum, there is a wealth of technical info on the E6750, as well as others. I can guarantee you one thing, my cpu is nowhere near a true 70c under load....

The GO stepping cpus, which the E6750 is, are different than previous models and have been pretty much shown that current monitoring software is not correctly interpreting the DTS, by a factor of 15c. 100 - 15 = 85c. Bingo.

Based on 65nm tech, they would have a lower tjunction max than the 90nm models, which also supports the data that tjunction max is 85c....too bad Intel won't release the pertinent data for all the C2D's and stop all this techie goofiness :)

Here ya go.....

"Thermal Specification: The thermal specification shown is the maximum case temperature at the maximum Thermal Design Power (TDP) value for that processor. It is measured at the geometric center on the topside of the processor integrated heat spreader. For processors without integrated heat spreaders such as mobile processors, the thermal specification is referred to as the junction temperature (Tj). The maximum junction temperature is defined by an activation of the processor Intel® Thermal Monitor. The Intel Thermal Monitor’s automatic mode is used to indicate that the maximum TJ has been reached. "

Since Tcase for this cpu is 72c, measured at the top of the heat spreader, it stands to reason that the Tjunction max, measured via the internal DTS, is 85c. That falls right in line with other models as well and provides support for the E6750 being an 85c max tjunction cpu, which is why I trust and believe my delta to tjunction temps are accurate. I have a good cooler, it has a good TIM interface, good case venting, and is rock solid.

http://processorfinder.intel.com/details.aspx?sSpec=SLA9V#
 
Check out extremesystems forum, there is a wealth of technical info on the E6750, as well as others. I can guarantee you one thing, my cpu is nowhere near a true 70c under load....

The GO stepping cpus, which the E6750 is, are different than previous models and have been pretty much shown that current monitoring software is not correctly interpreting the DTS, by a factor of 15c. 100 - 15 = 85c. Bingo.

Based on 65nm tech, they would have a lower tjunction max than the 90nm models, which also supports the data that tjunction max is 85c....too bad Intel won't release the pertinent data for all the C2D's and stop all this techie goofiness :)

I'll be sure to check out that forum, thanks for the heads up.
 
Ouch while we're at these puny OCs and talking about temperatures, have you just seen new Wolfdales at 1.5V/4.37GHz? And more or less acceptable 70C under Prime! And that's with Zalman 9700, nothing more :eek:

Looks like we'll all be turning our heads to these babies as soon as they go down in price a bit :D Than i won't bother with same 70C on just 3Ghz :p
 
so correct way to interpret the temp on my e6750 would be 85 - 33 = 52c idles around 30c

which is what coretemp shows, the tjunction is reported wrong as 100 -49 = 51c
which would still be the correct temp just wrong info.

im pretty sure its not that hot.
anyway i used a gemini II free after rebate , with a thermalright 120 backplate , and 2 yate loon dsls running at 1500 rpm, its a shame this p5wdg2-ws wont go over 400 but i bought it for the pci-x anyway.

 
im pretty sure its not that hot.
anyway i used a gemini II free after rebate , with a thermalright 120 backplate , and 2 yate loon dsls running at 1500 rpm

Man, no matter how you put it, you're having load temps not far from where I sitt at idle :D YOu are one that doesn't have to worry a bit, just like Violator ;)
 
so correct way to interpret the temp on my e6750 would be 85 - 33 = 52c idles around 30c which is what coretemp shows, the tjunction is reported wrong as 100 -49 = 51c which would still be the correct temp just wrong info.

I had a post about Tj Max values in the Core Temp forums and the mod there seemed to agree that the Tjunction Max seemed 'off' in some rare cases.
http://www.hardforum.com/showpost.php?p=1031906867&postcount=7

I suggest you go by the more accurate Delta to Tjunction Max value as a rule of thumb if you are unsure of the Tjunction Max value of your C2D. Keep the Delta value between 20'C to 25'C.
 
Back
Top