Q9450 temps?

[Fallen Kell]

Ok, so I didn't keep up on all the 45nm core temp issues and what not, sorry (I was too busy dealing with the fact that I had a Gigabyte 680i motherboard that wouldn't even run a 1333 Quad, let alone a 45nm).

Anyway, I picked up a Q9450 for my 780i board system and just installed it. What temps should I be seeing without an OC, and what program gets them correct (or is there an offset that I have to do)?

With speedfan, it is saying 32C... With Core Temp 0.97.1, it is saying 57-58 for 2 of the cores and 52-54 for the other 2. Which is correct (or are they both wrong)? Oh, I have a Tuniq Tower with a S-Flex-F on it with the fan speed being controlled by the CPU fan header and the BIOS.

 

[bryanW1995]

I have a tuniq w/ stock fan, ip35 pro, antec 900 with all fans on med except top fan on high, and an X3350. Are those temps you show idle or load? my idle temps at 3.6 are pretty similar to what you're showing in coretemp, I think they're 7-8c lower at stock. my load temps are obviously much higher...I'm currently testing at 3.704 however so I can't go check idle atm

I wouldn't bet on speedfan, though there is mounting evidence that coretemp incorrectly reports tjunction as 105c instead of 95-100c. realtemp uses the same sensor as coretemp and is consistently 10c lower b/c it uses a tjunction of 95c. I've actually given up trying to find the "true" answer to this, I just look at delta to tjunction now and try to keep my cpu as far away from that as possible!

 

[Fallen Kell]

Those are idle... However, as I said, I am using the motherboard's built-in CPU fan speed controller, which is only running the fan at 10% power when temps are 30C or less according to the motherboard monitor, which seems to be what speedfan says, since watching the RPM's on the fan show it at 400 RPM, which I think is the low limit as to what it can detect... I will disable the fan speed control and just have it run at full speed and see what the temps say.

My main concern was that I didn't apply the TIM correctly. I used MX-2 for the first time, and it was a lot more "fluid" than MX-1 (well, just about ANYTHING is more "fluid" than MX-1 was, that stuff could have been used to spackle wall). I put a little too much on at first and then wiped away a lot (I used a Glad plastic bag which I turned in-side out and placed my hand in it to act like a glove), and in the process evenly spread it over the surface. The tuniq has a visible machining ridges on the bottom contact area, which I had debated lapping before. But I had great temps on the E6600 that I have used it on in the past, so didn't bother (I cleaned it with the Artic Clean stuff before this install).

 

[marrr]

At stock settings, im getting 42 39 48 48 with real temp 2.41. I got a true with a 2500 rpm fan. I just wish my sensors werent stuck.

 

[bryanW1995]

Originally posted by: Fallen Kell
Those are idle... However, as I said, I am using the motherboard's built-in CPU fan speed controller, which is only running the fan at 10% power when temps are 30C or less according to the motherboard monitor, which seems to be what speedfan says, since watching the RPM's on the fan show it at 400 RPM, which I think is the low limit as to what it can detect... I will disable the fan speed control and just have it run at full speed and see what the temps say.

My main concern was that I didn't apply the TIM correctly. I used MX-2 for the first time, and it was a lot more "fluid" than MX-1 (well, just about ANYTHING is more "fluid" than MX-1 was, that stuff could have been used to spackle wall). I put a little too much on at first and then wiped away a lot (I used a Glad plastic bag which I turned in-side out and placed my hand in it to act like a glove), and in the process evenly spread it over the surface. The tuniq has a visible machining ridges on the bottom contact area, which I had debated lapping before. But I had great temps on the E6600 that I have used it on in the past, so didn't bother (I cleaned it with the Artic Clean stuff before this install).

my tuniq has no visible machining. I would probably lap it if I were you. You have just found out the hard way that there's a world of difference between a c2d and c2Q.

 

[Fallen Kell]

I just removed the BIOS fan speed controller and set the fan to 100%. I also downloaded HWMonitor (which was listed in the stickied overclock thread). My CPUTIN reads 27-28C, and my cores are reading 45-52. I should probably still lap the Tuniq...

 

[Fallen Kell]

I just ran load tests (Prime95 with 4 threads), and my max load CPUTIN was 44C, and on the cores it was 65C on three of them and 69C on the 4th. I think I see lapping in the future.

 

[bryanW1995]

I'm at 1.35vcore bios/1.29 load @3.6 on my x3350. Running small fft's in a room that is typically about 26c ambient (or sometimes higher ) I get 77 max on core 1 and 70-73 on the other 3 cores. Oh, actually, that was testing at 3.704 @1.32 load last night, I didn't get over 75 on core 1 and was 69-71 on the other 3 at 3.6. I'm planning to reapply thermal paste and probably lap as well, you definitely should look into that.

 

[Fallen Kell]

Yeah. I just ordered some sand paper sets (250grit to 2500grit). What sucks is that I will now have to go through the same pain that I dealt with installing this board again in my case with the heatsink attached. The real pain is that the location of the AUX CPU power is such that I have to attach them before I put the board in, and as a result I can't easily get to a screw hole in the bottom right corner of the board (which is the only screw in that entire section, and it is closest to the heatsink, so it NEEDS to be screwed down there). I had to rig up a telescoping screwdriver, which I then had to mount an old hard drive magnet to the screwdriver so that it would hold the screw on the tip while I tried to get it to line up with the hole... took me 25 minutes just to get that scew in last time, and I wonder how long it is going to take me to get it out....