o good choice's :up:
hope this works with xp:D
o good choice's :up:
hope this works with xp:D
wow this could not have come at a better time, thanks guys will be putting this to good use.
Why does this work with the 4870's, the 4870x2's but not the 4850? Does it have something to do with the ddr5?
Now if only CCC wasnt locked and Rivatuner supported multiple cores for the X2 in WinXP.. :(
Working on my gtx280 :D
thanks STEvil
that was my worry would it volt both gpu's on my 3870X2 or just 1 :(
http://forums.guru3d.com/showthread....36#post3006736
Toss some support this way :D
Back to 800/950 for now... dreams of 850/950+ dashed! :(
you got it :up:
Sooo, what about memory voltage, wouldn't that be controlled by the same controller, i.e. upping = possible?
nice info thanks for the directions I'm gonna attempt this tomorrow night ...
Seems it works on my reference 4870X2 (PowerColor).
1 issue, I can only get the different voltage to work for GPU 0, not for GPU 1. ANyone with a 4870X2 help me out?
Stock 3D is 1,26V which I have identified to be found at: RivaTuner.exe /ri1,70,18
I wanted 1,4V (hex 4C) so I typed:
RivaTuner.exe /sd0 /wi1,70,18,4C /sd2 /wi1,70,18.4C
Like I said, only applies to GPU 0, GPU 1 stays at 1,06V and is being raised to 1,26V when I enter 3DMark while GPU 0 permanently stays at 1,4V. Any ideas?
Edit: Just disabled OC within Rivatuner and now it works.. gives both Cores the same voltage.
Also thanks for your bat file Fornwagon, it's spot-on with my setup (Same program path, same bus location) so I can use it w/o any mods :D
One last problem: When I OC within rivatuner, it only OCs core 0. Core 1 stays at the factory 750 Mhz. If I OC with CCC to 800, both cores get 800Mhz. So what can I do? I'd like to OC further with CCC, can I do a mod so it'll allow me to go higher than 800Mhz core? Or should rivatuner be working on both cores? It iesn't for me... using latest 2.22 ofc with Catalyst 8.12
doesn't the 'enableclockcontrolcloning' option (poweruser tab> rivatuner/overclocking/global) work in XP x64?
jcool you can use amdgpuclock tool with this patch ... http://www.nordichardware.com/skrive...16/ToolLib.zip
and set what ever clocks you want first BN:3,DN:0 click set clocks ( it will say it didnt work but forgrt that it does) then on dropdown meneu select BN:4,DN:0 then put in the clock you want same as the first ones;)) click set clocks ... monitor will go black or a sec. then tell you it didnt work ( again it does lol) you can check if you want with gpuz it will show your new clocks :)
dont use riva tuner for clocking only for fan speeds
now tell me that the batch fornowagain really does set voltage to both gpu's ! really ? how to disable oc on riva ..( or do you mean just dont use it leave the clocks stock thier) ?
your card is a x2 card right ? not just 2 cards in cf right ?
if so were in bizzness :)
Sapphire 4870 Blue PCB
http://i20.photobucket.com/albums/b2.../Dibujo-91.png
(The default is 1.2625v)
C:\RivaTuner\RivaTuner.exe /wi1,70,18,44
damn it! Even at 1.45v, my 4870X2 refuses to be stable at 825MHz. Temps look fine, but no go. :mad:
Makubex does this work on both gpu's ? both have 1.45 v ? not just gpu 1 ?
Bah, I still get artifacts in ATI Tools' Artifact scanner @ 820MHz even at 1.52v. :(
I guess I'm sticking with good ol' stable 800MHz. :eh:
Thanks, will try that sometime. However I got Rivatuner to OC both cores, you have to select the other GPU in the dropdown menu and adjust the slider for both :D
To answer your other questions, yes I got a 4870X2 all right, and yes it works on both GPUs, according to Rivatuner monitoring.
However, I am not seeing any gains from this. I can run 830 Mhz core 3Dmark and games at stock voltage, but not even 1,45V will prevent the 2nd 06 test (firefly forest) to lock up within 10 seconds. I did run at stock and raised vgpu and noticed increased current ratings in rivatuner, so it should work. However, it doesn't net any gains for me. Can't be the temps, all watercooled with a max of 36C registered. :(
And yes, I did the whole calculation process for myself to understand what I was doing, but fornowagon's batch file worked for me without changing anything (same I2C bus address, same rivatuner version, same rivatuner path)
That's exactly what it is, both 4870X2 limit out at 1.371v. No matter what voltage you set (e.g. +1.4v)
http://i180.photobucket.com/albums/x...n/DSCF3124.jpg
Makes sense, I could only see an increasing VR current up to the 1,4V setting.
Still the whole affair makes no sense, why would 1,37V give the same core OC as 1,26V? It should be at least 20-40mhz higher but it isnt.
Why did you write the batch to go up until >1,6V if it doesn't work anyways? :D
You may just be at the limits of the shaders. Both core are definitely stepping up and very close to the register values until 1.37v.
Wonder if its the same and what the OVP is on 260/280?