[ELW]

I recently installed a passive Ninja with Thermalright bolt-thru kit on an E6600 based system.

Do the following temperature readings indicate that the system is getting the expected cooling from the Ninja?

Is it unusual that this system's Core temp readings are lower than the CPU temp readings?

System:

Case: Antec Solo mini tower
Case fan: 1 x Nexus 120mm @ ~740 RPM.
Motherboard: Asus P5W64 WS Pro
Processor: Intel Core 2 Duo E6600
Processor heatsink: Scythe Ninja Plus Rev B. Passively configured -- no fan.
Heatsink bolt kit: Thermalright LGA775 Bolt-Thru-Kit
Thermal material remover: ArctiClean Surface Purifier
Thermal compound: Arctic Silver 5 Premium
Video card: 2 x XFX GeForce 7600GS 2x DVI fanless
Memory: Corsair 2x1GB
Hard drive: Western Digital Caviar 250GB
Optical drive: Samsung DVD±RW DL 20x SATA
Power supply: Corsair HX520W
Monitors: 3 x Dell UltraSharp 2407WFP-HC 24-inch
No other fans, except of course for the one inside the power supply.

Test applications:

2 x CPUburn (two concurrent instances)
Windows Task Manager Performance screen
SpeedFan 4.33
Asus Probe (for validation)
CoreTemp 0.96 (for validation)

Measurements:

Ambient room temperature: Roughly 13C
Load: Idle
System: 37C
CPU: 32C
HD0: 36C
Core 0: 27C
Core 1: 27C

Ambient room temperature: Roughly 18C
Load: Idle
System: 39C
CPU: 34C
HD0: 38C
Core 0: 29C
Core 1: 29C

Ambient room temperature: Roughly 32C
I built a hot fire in the fireplace to create a high ambient temp for testing. I wanted to simulate a worst case environment.
Load: Idle
System: 46C
CPU: 41C
HD0: 42C
Core 0: 36C
Core 1: 36C

Ambient room temperature: Roughly 32C
Load: Full 100% CPU for an hour.
System: 47C
CPU: 62C
HD0: 43C
Core 0: 60C
Core 1: 59C

[Ryan]

are you using coretemp for your core temp readings, speedfan has had some difficulty with offsets in the past, youre real readings will probably be 15c higher as its impossible to have core temp lower than the cpu temp.

strange, temps look fine to me though, very impressive considering there is no fan!

[ELW]

Yes, I used CoreTemp to validate SpeedFan's Core0 and Core1 temp readings.

I used Asus Probe to validate SpeedFan's CPU temp and fan speed readings.

The temp readings almost always matched exactly. Occasionally I would see a 1 degree C difference between the tools. My guess is that the temporary difference was caused a sampling lag due to running the CPU at 100%.

Here is a snapshot that I took at an intermediate point during the load test in the high ambient temp room. SpeedFan and CoreTemp readings match.

[Shinigami]

Hi,

It looks fine, and I would say it is really great becasue it is "fanless". I would not complain about the temperature you have. Really safe even 32°C ambient.

[ELW]

OK, thanks for the feedback.

I'm not sure why the "CPU" temperature readings are higher than the Core readings.

My best guess is that the "CPU" readings are possibly not actually CPU but something else?

Maybe the Intel 975X chipset reports temps a little differently?

[Shinigami]

Well, I assume that they use a calculation to get close to the temperature of the Cores by that, as the sensor might be bit more far or influenced by other hot components.

It all depends in the Mobo and It's BIOS.