Graphics HD5970 'Fixed all problems with latest official BIOS' (Added latest vBIOS in post #1)

i have a waterblock on it,

here you have screenshot of aticatalyst, you can see it works, maybe i have a stupid card, hehe

http://img217.imageshack.us/img217/3179/hd5970.jpg

 

As you have RBE, why don't you open RBE and load one of the bio and see which one has Fan control high lighted.
The one with fan control high lighted is the Master vbios.

I would check devices manager and see if you have 2 x hd5900 in there with no errors.

 

here in device manager, its danish language,but you can see, it show 2x5970 and mine 8800gt



http://img824.imageshack.us/img824/8321/5970h.jpg

the first file 597ACNB1.bin http://img204.imageshack.us/img204/2412/masters.jpg

597ACNB0.bin http://img842.imageshack.us/img842/4857/slave.jpg

 

Ok, that seems to be correct.
You can retry the vbios in post #1 again to see if it fixes any problems you may have had.

 

i tried again, no errors, nothing. After flashing, reboot and i could hear the beep, could see on my lcd screen poster, that it went through vbios chek, and booted in to the windows, but no picture on the screen.. Flashed it back to orginal bios, and its up and running again. Dont know whats the problem

Here is some pic, of the flash list, so you can se, if everything looks ok

pic #1 http://img842.imageshack.us/img842/2934/iphoneslike006.jpg

pic #2 http://img801.imageshack.us/img801/5557/iphoneslike007.jpg

 

If you look at OLD PRODUCT NAME and NEW PRODUCT NAME, you are installing the Slave on the Master and the master on the slave.
Try again.

Edit; I don't get it, try it this way,

atiflash -p 0 C0001OCS.AC1 -f

atiflash -p 1 C0001OCM.AC1 -f

 

Have you uninstalled all ATI CCC and rebooted, then used driver sweeper and then rebooted, before flashing the card?

If not, then thats your problem, as your XMP Profiles do not match up with the vbios on the GPU's

 

never minde, i thought i was booting on mine HD5970, than i saw i did not switch the dvi cabel between 8800gt and HD5970,

When i flash it, i can hear the beep, but no pic on the screen,

flashing it back to orginal files, than no problems

 

Hi there, I only got an account just to say thanks to the OP (true_gamer)
THANKS DUDE!!!!!!!!!

I put up with that stupid cold start issue since last December
tried a few different BIOS's, none seemed to help
great find, and thanks again

also that seemed to help out my overclock on my DDR5

anyway thanks again....
btw have a look at my RIG u helped fix:

http://www.hardwarecanucks.com/forum/case-mods-worklog-gallery/

 

HD 5970 same issues, different solution!!!

Hi 5970'ers, let me give you my experience struggling with this 5970 + waterblock hell and how I found out the issue and finally resolved it.

Almost two months ago I bought a brand new Sapphire 5970 OC Edition to join my actual rig. A Gigabyte EP45-UD4P (stock cooler) + Q9550 (thermaltake WB, I know it's lame) + 8Gb Patriot 800Mhz (heatspreaders) + Gigabyte 8800GTS (Danger Den Ione WB), all the waterblocks powered by a Koolance Exos-2 unit.

Since I have no WB for the 5970 and the space inside the case does not allow me to run the two cards in that configuration, I took out the 8800 and put the radeon alone. Just the CPU was watercooled then. All was ok, playing a lot Crysis Warhead, SF IV, Metro 2033 and so on. MSI Afterburner + Kombustor were ok, and Unigine works awesome. My first time tesselation experience... . All was running over a win 7 x64 platform.

I ordered then a Danger Den WB for the radeon. When it arrives I happily mount it on the card, also thought that it would be a good time to bring back the 8800Gts with it's WB. But since I had a CoolerMaster iGreen 660W PSU, I also bought a Corsair 950TX PSU, so I replaced the PSU and added the 8800 GTS to the rig. I installed the 10.6 catalyst drivers and the 197.xx Geforce drivers plus the compatible hybrid physx mod... All was good, but just for a few days.

About three or four days since the above installed components I had been experiencing suddenly frezzes while playing. I thought that the stock MB cooler wasn't enough for the two gfx installed cards (Don't know if its related, but that is what I thought). I'll do some testing and put a fan over the heatsink. It worked for some time but the lock ups appeared again and again.

It was time for a radical change. I ordered an entirely new watercooling system since I thought the problem was chipset temperature. It also give me a good reason to convince myself to change that old equipment. The new parts were conformed by a Koolance Exos-2.5 unit (since the original one seems to have a non working pump), wb for the southbridge, wb for the northbridge, two wb for the vregulators and a new wb for the cpu. The gfx cards had already their own danger den wb's. Also I orderder an extra Laing DDC pump (the same the exos-2 sports) and the bitspower miniwater tank. The idea was to route the variable current the exos provide to its own pump the the new one to double the power pump into the loop. I also planned to install an extra fillport attached to the bitspower miniwater tank (the koolance unit has its own) and a drainport in some place. Finally a flowmeter will let me realize that all was working ok.

All the parts arrived in good shape. I needed a weekend to set up all the things, including the fillport, drainport and the second reservoir + pump and flowmeter. The loop was Exos-2.5 (rad -> pump) -> cpu -> northbridge -> flowmeter -> 8800Gts (here was attached the drainport) -> reservoir (here was attached the 2nd fillport) -> 2nd pump -> southbridge -> 5970 radeon -> vregulator -> vregulator, all mounted on a 1/2" feser UV blue tubing. Thanks to the second fillport, the loop filling was absolutely straight. I started the loop with no pc parts connected and worked really great!!!

I did the mandatory 24 hour leak test, passed ok. Well, there was time for serious gaming!... I then connected all the PC parts, computer booted correctly, but when I want to play something I got a hard lock up... Reset, rebooted... freeze again. No matter what I did, the PC freezes when I start a 3D game.

That was the begining of the nightmare...!

Well, all seems to be a driver issue. I uninstalled the drivers, reinstalled, driversweepered it, reinstalled again... Nothing.

Well, good old internet will have the answer I thought. I found this epic true_gamer thread which had a lot of information related to the same problems I had experiencing. After reading all the thread and a lot of other forums. I went for the vbios flash solution. I did it with the latest official bios True_gamer posted (v126). Nothing happens. I had the same problems. To make the story short, I tried all the below things after the vbios flash:

- Reinstalling drivers (10.2, 10.3, 10.4, 10.5 and 10.6) with/without geforce drivers.
- Enable/disable ULPS
- Enable/disable catalyst A.I. in standard/advanced mode
- Taking way the 8800gts (since the problems started when I put the card together with the radeon)
- Run the system with the old PSU (since the problems started when I change the PSU also)
- Fresh install of Win 7 x86 with all the above combinations.
- Reflashing the vbios with the overclocked true_gamer version.
- Reflashing with my own RBE-clocked vbios version.
- The two layer thermalpad over the chip in the upper side of the card (mentioned in some forum).
- Uninstall / reinstall the waterblock to be sure all the heating parts were in good contact with the block. All seems ok in every uninstall/reinstall (note that this means draining / refilling the loop!)

Nothing works for me, we are talking about at least three weeks of testing. I thought I had a hardware incompatibility. Then I pull out the radeon, put the stock cooler and tried in a fresh win xp install over a spare MB card I managed to get from my job place (I didn't want to test on my rig's MB since I would have to remove the southbride wb cooler and the pump + waterblock)...

And yes, as you are already guessing, it worked like a charm. Then tried in a fresh win 7 x86 install. Worked like a charm again. The conclusion was the 5970 was obviously ok. So, the EP45-UD4P MB could be faulty. So I had to do what I didn't want. I uninstalled the southbridge wb and the pump/reservoir, uninstalled the 5970 wb, reinstalled the stock cooler. I had to do a quick test because all the remaining wb stayed there but since I couldn't filled the loop cause it was not closed I had little time to test until the heat goes up.

After setting up the new test bed and boot up the system (with my thumb finger over the southbridge!), MSI Kombustor works flawlessly! Just to be sure I installed again the dangerden WB. Kombustor not even starts!... Freeze and hard lock. Had to turn of PC.

All indicates that there is a WB + Card incompatibility. Uninstalled the wblock and I began to look carefully the stock cooler and the WB. There was nothing strange, nothing different... just... aha! The cooper part of the stock cooler, the part that is in conctact with the two GPU's and the bridge chip has a transparent plastic layer. Now I looked into the card and If you do that you'll see a full row of little resistances just above the bridge chip. If you look closer you'll see they are just almost the same height of the bridge chip... This must be the culprit of this hellish freezing nightmare I thought. I put isolating tape over the WB to isolate that resistances, reinstall all... And now I'm a very happy 5970 radeon user!!! God damn it, what a troubleshoot session, probably the longest I have ever had!

I think that the first freeze just occurs because a material expansion because of the high temperature making short circuit between cooler and the resitances that were ok in the first place.

I hope this could help some of you. This is a very expensive card and there is a lot of guys RMAing their cards and maybe they are not faulty, just the same "incompatibility" between your wblocks and cards.

Regards,
Goblio

 

+REEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEP

 

I almost forgot...

Well, I almost forgot to thanks True_Gamer for his invaluable help getting the latest bios and pointing us into the right path to solve this annoying issue...

Thanks again dude!

 

I'm glad to have helped sort out your problems.
Lets just hope we don't have the same problem with the HD6000 series.

BTW your link doesn't take you to your RIG.

Cheers,

Simon.

 

I'm glad I have help you out
Great story there. I know it's a nightmare when you have your whole rig watercooled, as it takes so long to test different components to see what is faulty. Plus the dreaded thought of having to drain the loops to disconnect each component. What should be a 5 min job turns into a couple of hours. I have 3 loops in my PC and I am dreading when I upgrade my components.

Cheers,

Simon.

 

true_gamer first of all THANK YOU SO MUCH for sharing this with us.

I have the same problem you've mentioned, an ATI 5970 with an EK waterblock (with the extra thermal pad). In my case, I suffer from 2D artifacting on windows 7 desktop when GPU temp is below 36ºC.

Enabling a custom OC profile under catalyst 10.4 to fix 2D clocks at 400/900 does help a lot. But every now and then, when the water is still cold I have some artifacting, the card works perfectly after 2 minutes of furmark so it's definitely a cold bug. (EDIT: I've also tried other catalyst versions 10.5 and 10.6 to no effect)

I bought the card on November 2009 so I probably have an old BIOS version. My card is an ASUS EAH5970

I was wandering if there's a way to rising the 2D clocks by software on an ASUS board without the BIOS update you mentioned.

And if a BIOS update is the only way to go, how can I update to an ASUS Bios so I don't void my waranty? does anyone have an ASUS BIOS that can fix this?

thx!

 

finally, its up and running now. No cold bugs, no driver crashing, works great now. No need to raise 2d clocks to make it stable. Now i can run 3d application, with temperatures below 30C.

http://img340.imageshack.us/img340/3179/hd5970.jpg

so big thanks to TRUE_GAMER

btw aticatalyst 10.7 is out for download http://game.amd.com/us-en/drivers_catalyst.aspx?p=win7/windows-7-64bit

 

Hi,

You have already voided your warranty by adding a EK waterblock to your card so flashing with the sapphire vbios won't matter.

You can use the same vbios on your Asus card. I would save the original vbios so you always have the original.
There have been many people with HIS/Powercolor/Asus etc who have use the Sapphire vbios to flash there cards. With having the original vbios, you can always flash it back.

If you have not flashed a GPU before, I would recommend having a spare GFX card just in case the flash ever fails you.
As long as you do as I stated in post#1 you won't have a problem with flashing your card.

Cheers,

Simon.

 

Good to see you got it running in the end.

P.S thanks for the heads up on 10.7 being out. I'am downloading now.

 

XFX5970 BE and Cold Bug removal bios

Hi,

its a long time now I am following this thread and I would like to express a couple of questions to True_Gamer (Thank you for your contribution!) and Pepi-bih...

Pepi-bih:
Could you please post the exact commands and procedure you followed to flash the XFX5970BE with the new bios? I have the same card (no 2nd card available in case of mistake) and I really hesitate to compromise a wrong flashing procedure.

True_Gamer:

Could you please confirm that this new bios does NOT raise the 2d clocks? If I raise through CCC the 2d clocks the cold bug disappears, BUT I also have increased temps and of course increased power consumption when in idle.

Also, I have checked RBE and really cannot identify a difference between XFX 5970BE and Saphire bios parameters, except the slight increased default clocks of the saphire e.g. 1010mhz instead of 1000mhz. So what makes the difference with this bios... perhaps I could contact XFX and ask them to make a new bios including this difference.

Note: I have sent endless tickets to XFX about the cold bug and we ended up to the following solution (after 3 weeks of reasearch): send the card back to my dealer for a test. The thing is that the only reason/time I get the cold bug is when my room temp is 25C or below and start up my PC after a few hours. I do NOT have watercooling or overclocked my card. I just have a pretty fast SSD and my windows 7 64bit boot up in 14 secs resulting to the cold bug effect if the GPU is not preheated to above 35C. This conditions are not easy to be simulated in my dealers service room since they use traditional HDD and during the boot-up the card gets enough hot to override the cold bug temp threshold.

And of course since in the future I am looking forward to use watercooling or aerocooler I do not want to get stuck with this "cold bug" thing

Thanks a lot!

 

This topic helped me, I also successfully flashed my Asus card to the Sapphire bios.

One question remains though, True_gamer: Is it possible to edit the sapphire bug-free bios in such a way, that the card is still detected as an Asus card? With this new bios, I cannot use Asus Smartdoctor anymore, and thus I cannot use the Asus Voltage Tweak.
I tried changing the subvendor ID, but that didn't help me out much, certain configurations led to the drivers not even recognigzing the card.