• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

*** Official AMD Phenom II X6 1055T & 1090T Overclocking Thread ***

Angel Rex
I have done just that.
The CPU turned up before the Crosshair and being keen to try it out i temporarily gave it lodgings in a GB 770 rev 2

I had to flash the bios to the latest to get support for the X6.
It worked but there were issues.

1: The core voltage set itself to 1.3v on warm starts and 1.475v on cold starts
This made overclocking tricky.
CNQ seemed to kick in on warm starts but not cold ones, regardless of whether or not it was disabled.

Core voltage is set in bios relative to default, and not an abosolute value, so it is very dangerous to increase it.
Fine on a warm starts, but cold starts could seriously overvolt the CPU.

I haven't checked to see if there is a newer bios for it, i would like to think GB have fixed this issue.
Seriously check out if this is the case, before attempting any heroic clocking.


2: The Voltage regulators got very hot at high clocks.
No heatsink on the mosfets.
Not a problem maybe for gaming and general use, but no way was i going to do 12hr+ prime testing with it :)
 
Last edited:
Going to try out running the 1090t in an ASUS M4N72-E.
I don't expect it to be plain sailing like it is on the Crosshair 4

There is a newer bios for it so i will give it a try.
Ver 2005 was truly awful heres hoping 2201 is better.

Edit:
Bios still too buggy.
CPU multi's all over the place.
To get 20x you need to set 23x and that only works on a warm boot.
Can clock using FSB with multi set to auto but it isn't very stable.
Maybe will revisit when another bios is out for it.
 
Last edited:
Cozman. i've just had a thought, TMPIN2 could be the voltage regulators temp, on our motherboard.

:) ... I did mention that a few pages back ... but not sure .. doesnt seem to matter what case fans I use or which direction they blow/suck the temps I'm reading stay fairly constant ... TMPIN1 = 49C-51C TMPIN2 = 58C-60C .. always about 10C difference

Cant say I have noticed any throtterling when TMPIN2 is very high or that its higher or lower with vcore or NB voltage changes .. definately a weird conclusion here .. lol

as long as theres no smell of burning or smoke rising I'm trying to ignore it completely :D
 
screenie022.jpg


time increases and GFlops dropping ... is that throttling ? or normal ... NB starving the cores maybe .. not seen it change like that before

might be within a margin of error or could be I was posting here while testing .. darn I need to be more scientific .. lol
 
Cozman
That drop you are seeing is throttling.
Looks like your bios is using tmpin2 for thermal control
 
Last edited:
strange ... only does it with LinX .. running IBT the time / GFlop readings get faster (time) / higher (GFlops)

its certainly not consistant and thats the first time Ive seen it in LinX too ..

guess its a waiting game for most of us until we get a final release bios as there seems to be numerous problems with the current boards and beta bios's .. then we can maybe get some real testing with consistant results
 
With IBT you choose standard, high etc.
These tests are longer and therefore generate more heat the higher the setting.
Standard is no good to get things toasty, high is better and if you have 8GB then go to the MAX.

LINX chooses a test based on max mem so it is probably testing bigger primes than IBT on high hence higher temps ?
 
just out of interest I have tested with NB at default 2000 and default voltage

temps are exactly the same ... maybe a 1C improvement ... and still sees the same speed / GFlop changes in LinX ... so its not NB Frequency related .. will try ram at default voltage and SPD settings .. just when I thought I had everything sweet :(
 
Core temp, the program reports the motherboards SB temperature..

At least.. I think it does.
When my SB temperature changes, so does the "core temp", they're 1C within each other.
 
Core temp, the program reports the motherboards SB temperature..

At least.. I think it does.
When my SB temperature changes, so does the "core temp", they're 1C within each other.

coretemp reports the same temp as AOD for me.

Are you using the latest version which was released yesterday? 0.99.6.1
 
OK so I have got slightly more scientific and run several tests .. though not that scientific obviously .. lol

CPU remained the same settings for each test - 1090T @ 4Ghz 1.425v (bios setting but cpu-z reads 1.44)
between each test there was a reboot and the same things loaded (cpu-z x2 , HWMonitor , Windows Task Manager performance page & LinX 0.6.4 64Bit)

I will let you draw your own conclusion but will give my opinion at the bottom of this post :)

The ram settings were set as = 1600Mhz @ 9-9-9-24/2T or 1333Mhz @ 7-7-7-20/2T

Throttle is measured by the time/GFlops output during 10 runs of LinX and seen as a rise in time/drop in GFlops

Temps are given as the max peak recorded using HWMonitor during the LinX 10 runs

NB -----Ram Settings-----TMPIN1-----TMPIN2-----TMPIN3----- Throttle

3000 [email protected] 34C 52C 66C Y
2600 [email protected] 34C 51C 64C Y
2000 [email protected] 34C 50C 63C Y
2000 [email protected] 34C 49C 62C N
2000 [email protected] 34C 49C 62C N
3000 [email protected] 34C 51C 65C N

screenie023.jpg


My Conclusion : ... the throttle in my case is caused by the ram or the IMC and not by any of the temps ... maybe a IMC Throttle or the ram is getting too toasty but is Corsair Dominator with the heat spreaders
 
1333 7-7-7 is FAR better than 1600 9-9-9

Have you thought of playing with the HT clock to get the ram a bit faster at the 7-7-7 timings?
 
apparently there is a bios bug that when you clock the NB the HT link clocks up as well .. I didnt realise that before and left the HT Link on 2600 for ages .. lol

but I locked it back down to default 2000 as most peeps said it wouldnt make much difference as there is plenty of bandwidth at the defaults

no idea if it would make any difference to performance or temps but might be something I will test in the next few days
 
Back
Top Bottom