AMD FX OC'ers Club

[AphexDreamer]

Which temp do I look at? The CPU temp or each individual core?

And what is the max?

 

[nt300]

You loook at both. The CPU temp should be the high that your CPU hit and the indiidual temps is the current temp. I can't wait to pick up myself a Piledriver or a Trinity

 

[erocker]

Which temp do I look at? The CPU temp or each individual core?

And what is the max?


http://img.techpowerup.org/120314/Which Temp.jpg

CPU, as the core temps are wrong like on most AMD CPU's. Core temps are most likely around 5c higher than the CPU temp.

 

[Daimus]

CPU, as the core temps are wrong like on most AMD CPU's. Core temps are most likely around 5c higher than the CPU temp.

I agree, but not in all cases the core temps are wrong. Under extreme load cores shows the real temp.

Let no one be confused by the strange figures of the lower limit, the outside was -10 or a bit warmer.

 

[Mathragh]

According to AMD, the CPU core temperatures should be accurate from 45 C and up, so if you're worried about the "CPU temperature" not going over 61 C, you should look at the core temperatures for that.

I currently have my bulldozer running at 4,0 GHz with 1,260V(after LLC), and its not getting warmer(core temp) then 37C under prime95(stable). However, as the core temp isnt accurate below 45C degrees, i actually dont have a clue about the exact temperature. only that its above room temperature and below 45C lol.

Also, does anyone know what is up with thet fact that my chip seems to be totally stable for normal use(gaming/benchmarking) when prime95 will still sometimes show errors? I had my chip running at 4,5Ghz with 1,39 V for quite some time when i just got it. After i succesfully used it for two days in a row at 4,5GHz with 1,39V I started running prime95 and it showed to be unstable(and only in prime95).

Edit: The max temperature AMD gives for core readings is 61C for 8 core processors, and 71C for the rest.

 

[Super XP]

Daimus, How are you running your FX 8120 @ 4.40GHz with stock vCore. unless it's set at Auto, where it would automatically increase/decrease based on load.

I'm running my FX 8120@ 4.40GHz but with small bump in vCore.

According to AMD, the CPU core temperatures should be accurate from 45 C and up, so if you're worried about the "CPU temperature" not going over 61 C, you should look at the core temperatures for that.

I currently have my bulldozer running at 4,0 GHz with 1,260V(after LLC), and its not getting warmer(core temp) then 37C under prime95(stable). However, as the core temp isnt accurate below 45C degrees, i actually dont have a clue about the exact temperature. only that its above room temperature and below 45C lol.

Also, does anyone know what is up with thet fact that my chip seems to be totally stable for normal use(gaming/benchmarking) when prime95 will still sometimes show errors? I had my chip running at 4,5Ghz with 1,39 V for quite some time when i just got it. After i succesfully used it for two days in a row at 4,5GHz with 1,39V I started running prime95 and it showed to be unstable(and only in prime95).

If it's showing errors via Prime95, then your OC is either not 100% stable and/or Prime95 does not fully recognize the Bulldozer Architecture.

If I were you, I would drop the CLOCK Speed to 4.40GHz, then test again. I believe this speed is the sweet spot running on all 4 modules (8-cores).

 

[Mathragh]

If I were you, I would drop the CLOCK Speed to 4.40GHz, then test again. I believe this speed is the sweet spot running on all 4 modules (8-cores).

Hehe, yeah, I was coming to the same conclusion The temperatures and voltage scaling gets really bad after 4,5Ghz, and at 4,4GHz things still stay nice and cool(sort of, for a bulldozer ).

As for the Prime95 errors, I'm getting "ILLEGAL HARDWARE SUMMOUT's". Never seen them before untill I got my bulldozer. At the same frequency and voltage, it would do OCCT linpack fine.

 

[Daimus]

Daimus, How are you running your FX 8120 @ 4.40GHz with stock vCore. unless it's set at Auto, where it would automatically increase/decrease based on load.

I'm running my FX 8120@ 4.40GHz but with small bump in vCore.

I meant "4.4GHz stock voltage" when I say that I did not bump in Vcore manual (my mobo hasn't Auto option). Of course, Vcore is a little bit increased, because of Load Line Calibration is Full or something else.
You're right Super XP, it's not quite correct on my part and I'll take the word "stock" off. Thanx for the tip

 

[Daimus]

The max temperature AMD gives for core readings is 61C for 8 core processors, and 71C for the rest.

Small overheating is not terrible, if you do not plan to use the CPU for 10 years.

 

[HossHuge]

Hey, AMD FX OC'ers Club,

I just updated my BIOS for my ASRock 970Extreme4 mobo and in the description for that BIOS (1.60) it says " Add Turbo UCC function for AM3+" Turbo UCC is what ASRock calls its unlocking program.

When I read that it sounds to me that it is allowing for Bulldozer unlocks.

Has anyone tried before or recently?

 

[Daimus]

Hi, HossHuge
Never heard of the BD unlock before. Asrok really means by the term UCC unlock hidden cores. You can try to be the first.

 

[Blackcrow]

Hi this is my FX 4100 with 1.38v

The same CPU did 4.5Ghz @ 1.328v complete estable

 

[Mathragh]

Very nice! Congrats on the OC man

 

[daimonass]

Nice temp... What cooler ?

 

[Blackcrow]

Coolermaster 212+

 

[eidairaman1]

Cm v6gt!

 

[nt300]

Did ASUS new bios for Crosshair V allow to disable 1 core per module yet? Or you are forced to disable the entire module? 1 core per module gains you better performance right.

 

[Mathragh]

Did ASUS new bios for Crosshair V allow to disable 1 core per module yet? Or you are forced to disable the entire module? 1 core per module gains you better performance right.

Only per thread, but certainly not overall.

 

[AphexDreamer]

My PC started randomly turning off.

I'd unplug and restart and ignore it.

Then it got worse and wouldn't turn back on, reset CMOS and this morning I had the Device Boot Red LED come on.

Left for work and when I came back to mess with it CPU LED light is on and that seems to be where it is stuck at now.

Also I kind of hit the CMOS button once while the PC was on. Could that have harmed anything?

About to try diff things.


EDIT hitting the MEMOK button with it off it powered up and gave a short quick Beeb but then a few seconds later a long one and leaves the Boot Device LED on.

EDIT: Witched GPU's same issue so I take it it is not the GPU.

 

[pantherx12]

Optimised my over clock.

Running 4.5ghz at a reasonable voltage maximum temperature is 58 degrees after 30 minutes of intelburn in : ]

Going water soon so perhaps expect a 5ghz 24/7 clock soon.

@Aphex that sucks man, I think your right about pressing the button whilst it was switched on.

 

[AphexDreamer]

Optimised my over clock.

Running 4.5ghz at a reasonable voltage maximum temperature is 58 degrees after 30 minutes of intelburn in : ]

Going water soon so perhaps expect a 5ghz 24/7 clock soon.

@Aphex that sucks man, I think your right about pressing the button whilst it was switched on.

Im good now though. I changed SATA Cables and left one stick of RAM and it booted. As I was taking out the RAM I noticed that they were paired incorrectly since I'm using two 1333 and two 1600 (mean to replace those). So I put them back in paired correctly.

I'm good now although I think I need to replace my RAM. Sometimes when I turn on the PC, I'll have to restart it and it'll say Overclock failed but again I think that is due to my RAM.

 

[pantherx12]

I'm good now although I think I need to replace my RAM. Sometimes when I turn on the PC, I'll have to restart it and it'll say Overclock failed but again I think that is due to my RAM.

Ahh I get that message every so often!

I think it's a problem with fx only takes a restart to sort out for me, so there isn't even a problem I'm guessing, just false positive.

 

[eidairaman1]

Ahh I get that message every so often!

I think it's a problem with fx only takes a restart to sort out for me, so there isn't even a problem I'm guessing, just false positive.

probably not registering the right clocks and proper voltages needed for FX

 

[AphexDreamer]

Man so I thought it was my stata cable but I still got random shut offs not even BSODs.

So I tried each RAM module one by one and each time it wouldn't get past the windows logo.

Till I got to the last 1333 module and I'm finally in. Running off of 2GB uhhhh...

Guess I'll go ahead and get me these then. SAMSUNG 8GB (2 x 4GB) 240-Pin DDR3 SDRAM DDR3 1600...

EDIT: Scratch that. I just noticed that the board put its own settings with my CPU at 1400mhz going to test again maybe its my cpu? ARghh..

 

[pantherx12]

Man so I thought it was my stata cable but I still got random shut offs not even BSODs.

So I tried each RAM module one by one and each time it wouldn't get past the windows logo.

Till I got to the last 1333 module and I'm finally in. Running off of 2GB uhhhh...

Guess I'll go ahead and get me these then. SAMSUNG 8GB (2 x 4GB) 240-Pin DDR3 SDRAM DDR3 1600...

EDIT: Scratch that. I just noticed that the board put its own settings with my CPU at 1400mhz going to test again maybe its my cpu? ARghh..

Weird that : / hopefully it's easily fixable.


Earlier on I posted I was getting a bigwater 760 plus ( I still am but will be selling it asap) but managed to score a pretty decent rig for £350 quid which includes a Kandalf LCS

Will be selling most of it and if I sell my heatsinks other WCs it easily covers the whole £350

Hoping it will cool this 8120 nicely : ]