Bios Mods -The Best BIOS Update and Modification Source

Okay, at barcode I had the 5870m back in the day and stock volts were 1.05 which capped around 750mhz-ish. When you flashed a nodded vbios you could set it to up to 1.15v which was great I personally got as high as 925 core 1270memory. But on stock paste it heated extremely quickly and would throttle around 95-105 celcius I can't remember. But staying around 1.1volt was a happy medium for all around gaming not breaking records... So my question is are you locked at exactly on the values .87max and or .92 max or can you do this? Can you set .90 or .89 and see if you can still get similar clocks but not throttle maybe its just too much juice for its design? Like if .92 can get a clevo to 860mhz can you get like 800-750 at .89 or .90? My thoughts are can you run those volts as a test and see if it only throttles on vantage when over a specific volt or has it been on any volts over stock .87? Thanks I was very curious.

Ahmed, you mean the M17xR3 or M17xR2 ? Because M17xR3 is quite similar design to M18x - the EC FW is very similar too.
So you haven't performed any register/code changes to the BIOS in order to remove the throttling?
And do you know which entries in ACPI define the GPU/throttling support?
Your work there means that the GPU throttle is controlled via EC (used BIOS with GPU support, but older EC FW), correct?

Yes , that`s right , i will explain now :-
1 ) the A04 bios doesn`t support 580M GPU while A07 bios and A08 both support 580M GPU .
2 ) the A04 bios hasn`t got throttle with its default VGA unlike the A07 and A08 .
3 ) Nobody think that this is a VBIOS mod as the only VBIOS module inside the A04 , A07 or A08 is for the SANDY BRIDGE DEFAULT GRAPHIC CONTROLLER DEVICE ( INTEGRATED DEVICE ) .
4 ) when you compare both ECFW for A04 and A07 or A04 and A08 ,,,,,,,,,, you will find that the EC FW inside A08 or A07 is heavily changed ( alot of codes are changed and replaced ),,,,,,,,,,,,,,,,,,while the difference between the A07 and A08 ECFW is only one byte
5 ) i knew that blind flash didn`t flash the or didn`t update the ECFW as it hasn`t got any setting for doing this..............it didn`t even has file like phlash.ini to define what will be flashed...........Blindflash only flash the bios not the ECFW .
6 ) As the A04 hasn`t got any throttle but didn`t support 580M GPU.................while the A07 OR A08 support the 580M GPU and has throttle...........They flashed A04 from windows ( updated ACFW to the A04 one ).....then blind flash to the A08 ( used A08 bios but not A08 ECFW ).
7 ) the changes inside DSDT was about adding support to the new GPU and putting the new device in the D1 mode ( while D0 mode means in operation , the D1 mode means not fully operated or controlled )........and another changes in MXM , SSDT , VZL ......was replaced in the A08

Thanks for the info.. I'm wondering how could the EC have access to GPU throttling, because via the interfaces it supports it cannot reach such device.. Maybe it could be realised from EC to BIOS and then to the GPU. Or maybe some dedicated GPIO controlled via EC and routed somewhere...... I don't know, maybe someone with better experience in GPU throttling (how it's implemented) could help..

Anyway, I'm analysing the Compal EC-Host interface exposed to control throttling, however the docs don't say if only CPU or GPU throttling can be controlled too. The EC FW code does what the doc says (changes fan speed) + modifies certain internal registers, which I haven't yet analysed what they do... Maybe it would be worth to try to send such commands and see what happens.. I would need to add this into HWiNFO.

---

EDIT:
I have checked some more information and indeed throttling can occur if power load exceeds AC brick limitations and it's maintained by the EC.
Now the question is if this occurs because the GTX580M really draws so much power..

Ahmed, check Compal info and EC RAM registers AEh (THROTTLING_INDEX), AFh (THROTTLING_SOURCE), A0h too. They seem to be reflected in the DSDT too. They might be just status bits (no control), but if they reflect real status we might use them as a starting point to decipher the throttling mechanism.
We need an EC dump made using RW-Everything Tool of 2 situations:
1. No Throttling
2. Throttling

What about 40h~4fh commands !
Many of these commands has access to the ECFW and disable throttle !!
This means throttle is controlled via EC .

---

Ok sir i will check this in few hours as I'm outside now

Ooookay.. I made a test build of HWiNFO64 and added new controls to switch GPU throttling on/off in the Fan Control dialog of sensors. I don't know if it will work...
barcode - please get this beta build:
www.hwinfo.com/beta/hw64_385_1365.zip
Run, open sensors, then go to Fan Control, switch the new GPU Throttling options, test it and let me know if/how it works !

That's interesting
I will wait the results .

Hey there, long time listener first time caller here.

Mumak, the throttle is still there. I believe it is only temperature related, at least at my end. it is enough to heat up any of the two cards to 80*C (it is always the primary card taking generally higher load) for the throttle to kick in and it does not matter if we are talking stock or OCed clocks. When OCed, the cards simply heat up faster -> easier to hit 80*C.

When the throttle kicks in, the cards downlock down to "idle clock", so 50/135/101 (gpu/mem/shader), for a few tens of seconds and then go back to function normally.

I just tried your anti-throttle build and when only the temperature on GPU0 hit 80*C the throttle kicked in.

So 0x29a the gpu1 doesn't throttle with hwinfo new function? But gpu0 still throttles at 80c?

Hello guys,I am a green hand to flashing bios,could any one please tell me how to flash the original A00 bios which gives by johnknss?when I try to flash it by insydeflash in win,it told me that no EC firmware detected and canceled the flashing process,so how could I to make it?thanks a lot!