Fixing a Bad BIOS Flash on the 4870

[Solus Corvus]

If I had a dead card that died without a flashed bios, I wouldn't flash it and risk losing the ability RMA it.

[Mechromancer]

Little more detail about our 4 dead 4870s:

1) Lived for just over 48 hours from arrival.
2) Were retail boxed / non-review / HIS brand.
3) Lived for 2 days of testing, heat levels ~ 90C @ load.
4) Died approx. 2 hours after enabling CrossfireX.
5) Died after 3 reboots each between XP 32, XP 64, and Vista x64.
6) No BIOS mods, no fan mods, SmartDoctor not used.

Background information:

1) Still have ~10 working 4870s, none put into CrossfireX yet, due to symptomatic failure of others. Waiting to see...
2) No problems with any 4850s in any configuration - Tri and Quad CrossfireX tested.


Looked back at the packing slip, and the HIS lot was received from the same origin... Distributor is thinking that a bad batch may have leaked, due to a large number of HIS already coming back. I do know that these are reference cards, but it is conceivable that a bad batch slipped into production.

Lastly, I am intrigued by one thing here - flashers and non-flashers seem to be reporting the same events, down to a gnat's ass. Same no-post, same LED, same symptoms. Perhaps the flash brought about something that can die, or has the potential to kick in. I'll leave it in the hands of you all to figure out, just wanted to add a few last details.

These are getting shipped out on Monday, and we'll be hoping for the best on the others that we have so far. Good luck to everyone.

I've read about HIS having issues on OCF too. Also I think I saw a post about a guy having problems with a Visiontek too.

I would like to know what motherboard you tested these cards in. The make & model of the board may have an affect too (just a guess).

[Marvin_The_Martian]

Can you provide OP with the batch number so if other come in with the same issue he will be able to check if it's indeed a batch issue and get the proper responce going as quick as possible? Just a thought.

[NKD]

same with my card, when it worked all foru leds flashed, now the D601 is the only one that does not flash, and yea they never stayed on when it was working either, they would only flash once when you turn on the system. As I said I can't rma the card even if I had a chance to do it I would not, but it is none of my business on what others do, because it is their card. I would most likely buy one next month again if I can't get it to work.

[W1zzard]

can someone please save a 4870 bios from a working card using winflash and post it?

[Kensek]

I have my original sapphire HD4870 BIOS but it was saved with AtiFlash 3.59.

http://www.keepandshare.com/doc/view.php?id=660440&da=y

I removed the file since it has been determined that AtiFlash and WinFlash are truncating the BIOS file.

[W1zzard]

thats fine, link please

[NKD]

I have my original Powercolor bios saved with GPUZ if that works let me know I can send that to you.

[W1zzard]

for this test save with winflash / atiflash only, so we can see if there is some kind of issue between gpu-z bios reading and ati's software

[Kensek]

I edited my earlier post with the link in it W1zzard.

[spajdr]

another dead card by not flashing bios
http://www.nvnews.net/vbulletin/showthread.php?t=115579

[NKD]

oh ok, I really didn't save with winflash, so I guess that is out of question, thanks to all of you for your help.

[W1zzard]

ok looks like the bios images are NOT 1 mbit as a poster on rage3d claims, so thats not the reason for the problems

[W1zzard]

can anyone with a borked card measure the voltages on the board? i marked the test spots for you



on your dmm put the black probe to ground (like case metal or the metal piece on the graphics card) and the red one to what you want to measure

yellow area and green area should be either mvddc and mvddq memory voltage (dont know which is one because i have no board). any of the pads in one color will work fine, it is safe to short between same colored pads

the blue area is vddc (gpu core voltage). as you can see two solder pads go together (white boxes on the pcb), one is ground, the other one will be vddc. test both, the one with the 0 volt reading is the gnd one. dont create shorts between those pads.

if you have questions ask before causing more damage

[Logos]

another dead card by not flashing bios
http://www.nvnews.net/vbulletin/showthread.php?t=115579

card is not dead, read the thread again, user has solved his problem

[W1zzard]

card is not dead, read the thread again, user has solved his problem

gogo XSers check your cables. lol. afk football, hope the great german nation pwns the spanish

[NKD]

card is not dead, read the thread again, user has solved his problem

that is right, his issue was totally different, it was wasn't a good idea to call it a dead card, now back to topic, cheers.

[NKD]

gogo XSers check your cables. lol. afk football, hope the great german nation pwns the spanish

LOL, I tried everyway I can imagine, the D601 led never turns on.

[Logos]

gogo XSers check your cables. lol

his card is not dead, that's a fact, period. Who said a particular found cable issue would give a solution to others, who? It just did matter to mention that it was not a new case of HD48xx failure, that's all.

[XCheater]

fortunately I go with 4850 first
anyway after I have collected my money, I'll go with X38 and a 4870 of necessary if there is no more dead 4870 anymore, or I'll grab another safer 4850

[Dave Baumann]

ok looks like the bios images are NOT 1 mbit as a poster on rage3d claims, so thats not the reason for the problems

The ROMs on these boards are 1Mb ROM's, used for the first time used on AMD/ATI PC desktop boards.

[GoldenTiger]

W1zzard, I have my HIS 4870 original BIOS saved (unmodified, from before my card died). I'll upload it somewhere in a minute and edit with link.

EDIT: http://files.filefront.com/his4870st.../fileinfo.html

EDIT2: File removed as it is truncated per this thread.

[Wesker]

Very nice!

I'm starting to have doubts as to the quality of this Qimonda GDDR5. Apparently the 8800GT had problems due to their memory as well. I've also read a bit about how GDDR5 works and its NOTHING like the previous generations of memory. It varies clockspeed, voltage, and other parameters automatically to resolve signaling issues. Screwing with the BIOS is probably messing with this function in some way.

GDDR5 Articles:
http://www.extremetech.com/article2/...2310000,00.asp
http://www.eetasia.com/ART_880050357...P_74d933d1.HTM

Read up on how "signal training" works for yourself and lets discuss it. I'm no memory expert so I won't pretend to know how it all works. It just seems like these BIOS mods are the culprit.

The 8800 GTS also had Qimoda chips, and I don't see anyone complaining about them. The 8800 GT was plagued by an inadequate VRM for its given specifications.

Further; Qimoda supplies GDDR chips for many other cards, all without any issue. That's beside the point that; to even suggest that engineers at either Qimoda or ATI/AMD didn't put the GDDR5 samples through rigorous testing is way off base.

[Dave Baumann]

W1zzard, I have my HIS 4870 original BIOS saved (unmodified, from before my card died). I'll upload it somewhere in a minute and edit with link.

EDIT: http://files.filefront.com/his4870st.../fileinfo.html

Whatever tool you used to save that with it is not saving to the correct size; the ROM image should be 128K.

I've just looked at all the BIOS's collected on W1zzard's download site, and again they are all incorrect sizes and the BIOS is truncated. W1zzard I would urge you to remove those BIOS's

[NKD]

hmmm, if it is indeed the wrong bios size, I am with you o that one DAVE, we should stop it from spreading in any and every way possible.