• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Speedfan latest version temp reporting, not accurate?

S

sram

[H]ard|Gawd
2FA
Joined
Jul 30, 2007
Messages
1,850
I decided to redo my overclock to see if I can achieve higher speeds, so I decided to use speedfan to monitor temps. I downloaded and installed latest version and adjusted the 15 degrees it is known to be off by. I then downloaded the latest versions of core temp and realtemp because temps seemed high with speedfan. Both Core temp and realTemp reported the exact the same temps for all my cores and they were less by 5 degrees from speedfan. why? Speedfan is now off by 10 degrees?


I wouldn't trust it because two other good programs report identical readings! Or am i missing something?

And while at it, how do I ensure my fans are running at 100% all the time?

Thanks.
 
The trouble is that the TjMax for the Q6600 is unknown. We are led to believe around 100C so programs differ. People like to think there is some one magic number. Temp sensors are embedded into the CPU circuity in several locations in each core and the highest is reported for each core but all areas of the CPU internal circuitry cannot be covered or the CPU would be nothing but temp sensors. Relax. As long a you are aware of the differences and what is being used at TjMax and its a reasonable value, you are good. Realtemp is my favorite and you may (I haven't checked lately) be able to adjust the TjMax in Speedfan to what Realtemp uses so they match. I know you can adjust Realtemp to match speedfan. Also each and every CPU has a different TjMax calibrated at the factory so it is very possible/probable that any two Q6600;s will have actual TjMax values that differ slightly. The "distance to TjMax" (whatever TjMax might be) is dead accurate due to the way the CPU reports its temp and should be the overriding value to watch. The more the better. This value should agree across programs that read the internal digital temp sensors in the CPU.


Under PC Health Status bios submenu

CPU Smart FAN Control_______ [ Disabled. Enabled (Default) ]

setting CPU Smart FAN Control to disabled should force the fans to 100% speed.
 
Thanks bill..........All clear now.

As for the fans, I did figure it out myself in the BIOS. I thought Speedfan has something to do with it though.

I wish I can reach your 3.37 GHz with 1.387v. I'm currently at 3.33 GHz and I had to give 1.408 to this greedy cpu of mine to be stable.

See this realtemp screenie for my idle temps:

 
sram said:
Thanks bill..........All clear now.

As for the fans, I did figure it out myself in the BIOS. I thought Speedfan has something to do with it though.

I wish I can reach your 3.37 GHz with 1.387v. I'm currently at 3.33 GHz and I had to give 1.408 to this greedy cpu of mine to be stable.

See this realtemp screenie for my idle temps:

Click to expand...

being that its a Q6600 the idle temps are meaningless.. only temps you should go by are your full load temps.. but 1.4v isnt all that much.. you still have plenty of room voltage wise since its a 65nm chip..
 
sirmonkey1985 said:
being that its a Q6600 the idle temps are meaningless.. only temps you should go by are your full load temps.. but 1.4v isnt all that much.. you still have plenty of room voltage wise since its a 65nm chip..
Click to expand...

Yeah I know. I just put it to give an idea of what kind of temps I'm getting.

Here are my load temps:



Uploaded with ImageShack.us

A bit high I think !

If I try to reach 3.4 with more voltage, I know I will hit 80. This is with 30 degrees ambient temp.

Thanks.
 
I think those are pretty good temps, just looking at the temps, The key is that you have over 30C of "headroom" before the CPU even thinks its hot and starts to throttle. Most of us like to keep that Delta to TjMax around 15-20 or better to provide a margin so you have 10C more you could play with before I would consider that CPU to be working hard. Fuck bragging about temps, (not saying you are) its about tweaking the system to give max performance on whatever you do and keep it stable but not overly stressed IMO.

Hint, the trick to my OC ended up being the FSB voltage mainly and some MCH voltage too. Higher CPU voltages are no help at all for me. Apparently the quad is quite a load on the buss. Dont know why, just is. On my P965 board which is a generation older than yours, its the memory subsystem that holds me back. My same board would do 475 FSB with an E6300 and 2x1 GB. Going to 2xGB made me reduce my OC and going to the Quad was a big hit in FSB (offset some by the higher multiplier than the 7x of the 6300) but bottom line until I hit the right combination of FSB and MCH voltage and accepted that I would get no OC out of the memory my OC sucked. (And still does compared to a lot of Q6600's). I managed to eek out 3.37 ish absolutely stable. But it was a long struggle of taking notes, changing one thing, a little bit, seeing over a day or so of almost constant use if the change helped/hurt/no effect, and then trying something else. Cooling the MCH/Northbridge is critical, glue a fan on it.

Just noticed the 30C ambient. Thats pretty hot room so you temps look even better taking that into consideration. You need to get into advanced tweaking of the FSB buss and the MCH, I dont think more Vcore is going to do much for you and even backing off a bit and retesting might be good. Been wrong before.
 
I'll put in my 2 cents since I just recently spent a couple hours overclocking my q6600. I can post up screen shots if needed after work.
I'm not trying to brag, but I must have got lucky with my chip.I got my clock to 3320. 415x8.

If I recall I only had to up the voltage to around 1.26v. (According to cpu-z) the bios is set to over 1.3. I have ram at 2.0v, the MCH at +.2 I think.
I modded my northbridge and added a cooler from a bfgtech 680sli board. Its around 70c

Its been folding for about 3 weeks straight with that OC stable. according to realtemp I'm at 73-78c between the cores.
I had it booting to windows at 3700, 415x9. 1.4v. But it was overheating like mad.
 
Cooling the MCH/Northbridge is critical, glue a fan on it.

Just noticed the 30C ambient. Thats pretty hot room so you temps look even better taking that into consideration. You need to get into advanced tweaking of the FSB buss and the MCH, I dont think more Vcore is going to do much for you and even backing off a bit and retesting might be good. Been wrong before.
Click to expand...

I'll put a fan on the MCH, thanks for the tip. I"m actually thinking of doing another mod to my case so that I can reduce overall temps. I currently have two intake fans and two exhaust fans in addition to the PSU fan. It is kinda of hard but I'll try.

Regarding the ambient temp, I'm living in one of the hottest areas of planet earth. And it is summer now, so 30 is considered good here!
 
nimbyfaygo said:
I'll put in my 2 cents since I just recently spent a couple hours overclocking my q6600. I can post up screen shots if needed after work.
I'm not trying to brag, but I must have got lucky with my chip.I got my clock to 3320. 415x8.

If I recall I only had to up the voltage to around 1.26v. (According to cpu-z) the bios is set to over 1.3. I have ram at 2.0v, the MCH at +.2 I think.
I modded my northbridge and added a cooler from a bfgtech 680sli board. Its around 70c

Its been folding for about 3 weeks straight with that OC stable. according to realtemp I'm at 73-78c between the cores.
I had it booting to windows at 3700, 415x9. 1.4v. But it was overheating like mad.
Click to expand...

I was running at 8*415 but with 1.376v as you can see from my sig. That's a really good voltage for 3.32 GHz. What program did you use to stress your cpu? Prime95 small FFTs ?
 
I run prime95 small FFT,s for a short term test. Just to make sure its not going to overheat or bluescreen

Then I'll run the Folding@home SMP client. Considering it loads the CPU to 100%, and I've had it running full time.
That's all the stress test I need.
 
sram said:
I was running at 8*415 but with 1.376v as you can see from my sig. That's a really good voltage for 3.32 GHz. What program did you use to stress your cpu? Prime95 small FFTs ?
Click to expand...
The best way to quickly stress-test your CPU is with IntelBurnTest or LinX (they both work the same way). 5 runs only takes a few minutes and will quickly give you an idea of whether an overclock is stable or not. Once you find something that looks stable, then you can do some longer-term stability testing.
 
Zero82z said:
The best way to quickly stress-test your CPU is with IntelBurnTest or LinX (they both work the same way). 5 runs only takes a few minutes and will quickly give you an idea of whether an overclock is stable or not. Once you find something that looks stable, then you can do some longer-term stability testing.
Click to expand...

You are right. I just found out about this. I noticed that this test is really really stressful and hard on cpus. I could pass OCCT and prime, but fail intelburn test
 
sram said:
I'll put a fan on the MCH, thanks for the tip. I"m actually thinking of doing another mod to my case so that I can reduce overall temps. I currently have two intake fans and two exhaust fans in addition to the PSU fan. It is kinda of hard but I'll try.

Regarding the ambient temp, I'm living in one of the hottest areas of planet earth. And it is summer now, so 30 is considered good here!
Click to expand...

Okay......I installed the fan. I glued it using super glue, and placed it so that it rests on the video card....(Just in case the glue doesn't live up to its name !)

Here is an image:



Uploaded with ImageShack.us


Guess I don't have to worry about the MCH getting hot now........What do you think?

I wish I can monitor it !
 
Speedfan temps has NEVER been accurate for me.. my core temps are always around 10c lower than real temp and core temp
 
ThumperSD said:
Speedfan temps has NEVER been accurate for me.. my core temps are always around 10c lower than real temp and core temp
Click to expand...


thats because the tjmax is set differently from real temp or coretemp.. real temp was the same way for a while.. 1 version would come out and the temps would be 5-10C lower then the version before.. next one comes out they are 15C higher then the last version.. constantly changed depending on the information they received..
 
You must log in or register to reply here.
Back
Top