• 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 ***

alexp999 said:
how is it impossible, you still havent linked or explained how/why its software "guessing".

Maybe you could get hold of your AMD contact your using and ask about the difference not just one to use?

Cheers. :)
Click to expand...
so a 20c room and idle core temp at 15c, is posable?????????

lol

come on, thats impossible


The K10 reports a temperature value that is relative to a certain predefined value, it doesn't report the actual processor temperature! So take that into consideration.
http://www.alcpu.com/CoreTemp/howitworks.html#AMD
Click to expand...
 
Last edited:
Hey alex dude ;)

I'm pretty sure gareth is right mate - it correlates with everything else I've read on the Phenom II's - There are no individual core sensors hence why its 'software generated'. There is only ONE cpu sensor on the AMD's.

For example - I used to have a Q6600 and you could see a good 5 degrees difference between all the cores - EVERYTHING i have seen on the phenom II's which generate individual core temps are exactly the same for all cores - this just isn't technically possible as you should inevitably have differences.

BUT - saying all that - The 10 degrees difference between the four programs I've mentioned might explain why I was able to hit 68Degrees C the other day when messing around because the CPU should have shutdown if I'd hit the thermal limit - not sure about this though.....

Anyways - I've dropped my CPU volts to 1.42v and set 2000/2000 for CPU-NB and HTT - I'm hitting tops of 54degrees c (using Asus probe and HWMonitor) - looks like my IMC is being strained too much to go higher without raising temps substantially - thats what one gets for trying to be clever with 4 x 2Gb sticks of DDR1600 I guess - lol
 
Last edited:
I will hazard a guess that with the UD5P defaulting to 1.475vcore now I wont have any stability problems with 4Ghz ... though temps are obviously likely to be higher in the long run ... so far they have remained the same within a degree or two

suppose I could undervolt back down but is 0.03vcore gonna make much difference if temps are that close ..

HWMonitor - AOD - and the rest .. take your pick they are all saying something different but AOD is reporting the lowest being 6-7C lower than the rest

now to think of a game I have I could run to stress the system @ 4Ghz .. well one I have installed already .. lol something tells me Tiger Woods Online wont do it :D
 
The difference in 10*C is not a calibration thing, its two differen readings

Fire up everest youll see what I mean. There is a "cpu" temperature and a "core" temperature.

And Gareth, Im not disputing the fact that the actual reading might be unrealistic, Im disputing that it is a different sensor for "cpu" and "core", i had a q6600 that read its cpu and core temps below ambient.

CPU thermal sensors are designed to work in reverese, I know they are not accurate idle, but I still believe there are two seperate sensors, in some shape or fashion. One for core, one for cpu.

The core sensor is the temp read for all cores, you see it in AOD all temps exactly same.
Even HWmonitor shows you both, scroll down, youll see a temp reading for TMPIN and the "core" temps.

The program core temp only shows the reading of the "core sensor"

PC Probe only shows the reading of the "cpu sensor"
 
cant vouch for stability yet ... but

Gigabyte 790FX UD5P
Multi @ 20X
RAM set manually to 1333 9-9-9-25 2T or it defaults to 1T leading to stability troubles
All Voltage set @ AUTO = vcore set to 1.475v but in reality gives 1.49v
C&Q & C1E Enabled but makes sweetFA difference so far
TMPIN1 recorded max using HWMonitor during ITB High stress 20 runs = 54C

gotta be one the easiest clock going just upping multi and manual ddr settings but will need to run prime95 for longer tomorrow to test

encoded a 1 hour TV episode from DVD to x264 and played a round of Tiger Woods online at same time temps got to 50C and encode was complete in 20mins (full screen RIP 4:3 old British TV proggie CRF@19)

so Im happy if the encodes go that speed in general as thats why I went X6 and hopefully Bulldozer on the same board next year

EDIT: should add case is Antec 900 all fans on low and H50 on cpu with default corsair 1 fan intake
 
Last edited:
well it's done 1 hour at 4ghz 1.4v , TMPIN1 48c, core temp 42c




tomorrow ill go for a 8hour+ run...

i've left the ram and the cpu-nb on auto untill i get the cpu clock fully stable.
 
Last edited:
Impressive mate.
I've tweaked my last settings and on my final prime blend run myself XD.
Clocking a 1055T is annoying, I wish I'd got a 1090T.
But if I get this stable, I've got a mint deal for my 1055T lol.
 
dandos23 said:
Would I be right in thinking that it looks like most of these chips (1090T) will reach 4ghz stable then?
Click to expand...

well Im bound to be proved wrong .. lol

but I dont think anyone has failed to hit 4Ghz yet its just a case of getting them as stable as possible on the lowest vcore and with lowest temps

:cool:
 
sorry gareth, regardless of what that quote says i still dont think you get my point. Where are the two sensor readings coming from, how are they calculated and why are they consistent across apps if its software guess work. Ill try contacting amd and asus, see if they can shed light on the matter
 
alexp999 said:
sorry gareth, regardless of what that quote says i still dont think you get my point. Where are the two sensor readings coming from, how are they calculated and why are they consistent across apps if its software guess work. Ill try contacting amd and asus, see if they can shed light on the matter
Click to expand...
just because u can see numbers in the core temp doesn't mean its hardware based.



look here at tmpin4. that goes up and down, do u think thats hardware based???

impossible.....

phenoms only have one sensor
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom