Sunday improvement drama - Bricked card

[nnay_th]

Hi guys

As winter is over in Europe, and temperatures begin to raise, I decided to undervolt my R9 280Xs from Asus.
I tested it on my smaller rig, which has only two of those cards, here is the story:

I followed the guide from cryptobadger (and ignored his warning to abort if there was no BIOS switch to recover in case of fail, silly me...)

Modified the VDDC from 1.2V to 1.087V then flashed the modified BIOSes to the cards.

Card 1 had no problem, hashing at same speed as before and cooler than before.
Card 2 had several problems: lower hashrate; or sick; or HW errors; or not hashing.

So I thought I would bump a little its voltage, to 1.093V. Did the same, modified the BIOS then flashed it back to the card.

Problems were less frequent BUT upon some reboots, it had the same problem as not hashing (0 Kh/s). I thought I would put back the original BIOS... BUT then begins the BIG trouble: ATIflash would not complete the BIOS flash and has the "ERROR 0FL01".

I now have a nice and expensive paperweight 

Has anyone been able to recover a bricked GPU knowing that R9280X-DC2T-3GD5 do not have the safety switch?

(morality: only undervolt cards with a BIOS switch)

[crazyates]

Did you try reflashing the original BIOS with the "-f" option?

[nnay_th]

Did you try reflashing the original BIOS with the "-f" option?

Hi crazyates,

Yes, actually I bricked the card when I tried to reflash the original BIOS, using:
atiflash -f -p 1 1.rom

That's when I got the error "ERROR 0FL01"

[NameTaken]

Even if it's not hashing, is it still recognized by Windows thus not bricked?

[nnay_th]

Well, as it turns out, I was able to reflash it, but not with the original BIOS. I'm wondering if it is because of this UEFI thing (I remember that I had a warning with VBE7 regarding UEFI being removed.

I flashed it with a modified version, with 1.1V, so the card is up and running again! I mean, more like up and walking as now the hashrate went from 750khs to 560khs on the revived card... any suggestion?

Thanks for your support guys!  

edit: the revived GPU gets "sick" after 2 minutes of mining, what does it mean?
edit 2: I put the revived GPU to 1.093, I now get a @#%load of HW errors
edit 3: I put it back to 1.2V (but not via the original BIOS), and it works fine now