Jump to content

In search of the bleeding edge....again


Bull Dog

Recommended Posts

Note: I originally posted this in a forum where not no savvy tech people browse. So please, I'm not trying to insult your intelligence.

 

It used to be I kept my 4400+ OCed to 2400Mhz @ 1.325v, that was good enough as it was just as fast as AMD's fastest X2, the 4800+ Well after AMD came out with a FX-60 (Dualcore proc running at 2.6Ghz), I just had to try and hit 2.6Ghz too.

 

A000947.jpg

 

So far I seem to be fine at 2600Mhz (260x10) with 1.45volts. What I'm AMAZED by in how cool the processor in running, granted I'm running it in a freezing cold basement, and at 2400Mhz it will only climb to about 43°C but still I'm left thinking about Intel's Smithfield jobs heck, even thier 65nm Pentium Extreme Edition 955 Presler and its a 65mn (vs 90nm) chip. Incidently, the FX-60 is margenally faster than a EE 955, and it(the FX-60) draws so much less juce just it aint even funny. (The 4800+ vs the EE 840 was even WORSE)

 

Woops, forgot to add that those temps are while running 2 instances of Prime95. Which so far after an hour it hasn't resulted in any erros. :D

 

Will leave it running during the night, wish me luck. :angel:

 

AMD: bring on the 65nm AM2's with thier DDR2 support.

Share this post


Link to post
Share on other sites

I have reason to believe that they are accurate. I've had the same processor in a different mobo and the temps read the same. Please PLEASE, keep in mind that my computer in located in a cool(bordering on COLD) basement.

 

Trying again: I set the voltage to 1.35v+110%=1.485 up from the 1.475v I had it at before. The this board undervolts about 0.025volts (according to the Bios) and 0.045 according to CPU-z

 

Edit: I'd take another screenshot but you'd see so difference as CPU-z shows the same voltage (it just isn't jumping between 1.424 and 1.440 anymore, although SG now shows 1.43v)

Share this post


Link to post
Share on other sites

The CPU temp is accurate to within 3 degrees (up or down, so 45c on software could be 42c to 48c in reality) ... This is accurate enough for most purposes, but not accurate enough for CPU to CPU comparisons (checking improvements from better cooling will be accurate to a fraction to a degree even if

 

Since the CPU temp probe is inside the CPU, testing in different motherboards is quite pointless.

Share this post


Link to post
Share on other sites

Excellent point Death909. 3 deg C is close enough for me. But hey I'll probably wind up with a thermal probe sometime. Right now I just don't have the cash for anything. There is one exception to that however. If the mobo's bios reads the CPU temp wrong, it isn't the CPU's fault anymore.

Share this post


Link to post
Share on other sites

I would guess those temps are accurate enough, Mine maxes out at around 27C and I can tell that since my radiator is on the outside of the case. The processor will only be a couple (2-3) C warmer than the radiator itelf and it BARELY feels warm at all when running maxed out.I shut off the fans to the rad at night and it'll be around 38C in the morning when it is obviously warm.

Share this post


Link to post
Share on other sites

The CPU temp is accurate to within 3 degrees (up or down, so 45c on software could be 42c to 48c in reality) ... This is accurate enough for most purposes, but not accurate enough for CPU to CPU comparisons (checking improvements from better cooling will be accurate to a fraction to a degree even if

 

Since the CPU temp probe is inside the CPU, testing in different motherboards is quite pointless.

actually this isn't the case. each motherboard manufacturer has thier own algorithm that calculates the average diode temp. otherwise you'd see the cpu diode temp fluctuating wildly up and down, just like on the dfi nf2 diode temp with mbm5 set to 1sec refresh; and the socket temp shows a more gradual temp change. the socket temp is the one that is using the algorithm and the diode temp is real time. this is on the dfi nf2. but it's true for all mobos. they all use thier own algorithm to calculate thier temps so they'll stay stable and gradually go up and down. take the MSI NEO2 PLat mobo, everyone and thier mother were bitching about how bad the temp monitoring was on that mobo, and now it's fixed with a bios update that included a new algorithm coding for temp monitoring. also, take the Abit 865PE and 875P mobos, everyone was screaming about how hot thier cpus were in those mobos from Abit, but placing thier chips in other mobos left them with a lower cpu temp. Abit chose to use the MTU temp diode inside the Northwood instead of what everyone else used, the standard general temp diode. but the single most vulnerable place to heat inside a cpu is in the MTU unit. so Abit never changed thier algorithm for thier temp monitoring and people just thought they were getting bad temp data.

 

TGM

Share this post


Link to post
Share on other sites

Currently i'm running at 2.7 and so far it is stable. I'm only using stock heatsink and fan. I'm running a 270X10 my voltage is sitting around 1.57Vcore and my temp is sitting around 40-45 load.

 

Although I acheived this through many hours of screwing up lol.

 

I hope this gives some of you some hope on getting yours up that extra little bit.

 

I have a thermalright XP-90c coming in the mail so look forward to it getting here and trying to up it a little more, maybe even get a 2.8 stable.

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...