Speedfan vs Realtemp

[Ryan111]

Which one is more accurate?

Why do they show different results?

 

[baked188]

Speedfan have set the TJmax at 5c higher than coretemp. I tend to follow coretemp but was wondering the same thing.

 

[PermaBanned]

Isn't it because RealTemp uses the sensor in the die, and most others use the sensor on the motherboard?

 

[Ryan111]

So which one do i follow?

I have always used speedfan, but if realtemp shows a more accurate result then...

 

[PermaBanned]

I use realtemp, but i think that the important thing is to use the same program across all your tests. Then you can get an accurate comparison. I would recommend RealTemp though, i love it's simple design and it's never let me down yet.

 

[Ryan111]

If realtemp show everything 5 degrees cooler that means i have allot more thermal headroom...

 

[Ryan111]

What is the maximum vcore you would use for a 45nm chip?

Im currently on 1.375 (1.344 after vdrop)

 

[ns400r]

If Speedfan is showing +5ºC and you know it is and are happy to use it, adjust the advanced settings and lower the figure by -5°C.

 

[Hughsy]

Mine showed roughly a -5c difference between both aswell.

 

[Ryan111]

I tried getting 3.6 out of my chip, thought it should be quite easy.

So I upped the FSB to 288, kept same volts and IBT failed on 2nd loop
Then upped the volts by 1 notch, failed 2nd loop again
Then finally upped the volts to 1.3875 i think it was... still failed IBT on 2nd loop. And it increased temps to 64 in realtemp.

Should the extra 200mhz require more than 2 notches of power?

 

[reflux]

I've always used CoreTemp although recently switched to RealTemp because I prefer the layout. Never liked SpeedFan.

 

[Jimbo Mahoney]

Realtemp is more accurate, assuming you calibrate it.

 

[Ryan111]

Calibrate it?

 

[Mr Paul]

Yes, calibrate it

 

[Rab]

speed fan 4.39 shows me same results as everest & real temp

 

[Mr Paul]

It all depends what TJMax these programmes use for the measurement. Coretemp and Realtemp will use the "correct" TJMax, speedfan does not for some CPU's.

 

[Ryan111]

Ohh okay then, thanks for clarifying

 

[El Jimben]

I tried getting 3.6 out of my chip, thought it should be quite easy.

So I upped the FSB to 288, kept same volts and IBT failed on 2nd loop
Then upped the volts by 1 notch, failed 2nd loop again
Then finally upped the volts to 1.3875 i think it was... still failed IBT on 2nd loop. And it increased temps to 64 in realtemp.

Should the extra 200mhz require more than 2 notches of power?

It varies from chip to chip really but I suspect it's not that, 3.8GHz should be possible as a couple of guys here have there E5200 @ 4GHz Did you try upping any of the other voltages such as the MCH and VTT?

 

[raptormonkey]

So, CoreTemp is the better program?

 

[PermaBanned]

It's a matter of opinion, but i'd say yes.