New Posts  All Forums:

Posts by PEJUman

Make sure you read the instructions carefully, you'll want to try the 4low.
tx12, I ran the verification command and got failures at x21 with both stock and modified bios. oddly the -p just said it is verified.I then flashed the _4low again and cold booted instead of restart. my Fury seems to still boot ok.In fact, I just completely shut down and rebooted prior to typing this message.NOTICE THAT EVEN REFLASHING THE STOCK BIOS STILL FAILED WITH x21 ERROR Is there a checksum in the bios?Is there more detailed manual on the atomtool I can look at? I...
just a thought, rx7 seems to OC'd his memory. rv8000, did you overclock yours as well? I was not brave enough to OC my memory and left it stock for now.
\So far mine still works using tx12 method from yesterday, both cold boot and hot restarts. Can you guys confirm that you're using 256KB sized bios, extracted via atiflash?I killed my first card using the wrong bios (128KB from GPU-z). The FURY X 128KB bioses don't really unlock but restarts fine. once I shut down, however, it failed the subsequent cold boot.
\Fixed my issue, turns out I had to restart the GPU detection. it's odd that windows didn't do it upon reboot and 3Dmark still works.I used the CRU's utility restart64 to force the driver to re-detect the GPU, and now it shows correctly as 3840 unified shaders.
Did you had to do something special to get GPU-z to show 3840? I rebooted and reran GPU-z, it shows 3584 with both 'all' and '4low' bioses. My firestrike ultra score improved from 3604 tp 3688 with the '4low' bioses.
Followed your instructions, thanks BTW.Used *4low.rom based on my cu info layout posted here earlier.My 3Dmark firestrike went from 3604 to 3688 @ 1000Mhz. No graphic corruption that I can see.Considering fury X @ 1050Mhz does 4000, and the score seems to linearly scale with clockspeed, I estimated 3700ish for 60/64 unlocked Fury.So my score makes sense.However GPU-z failed to show any shader increase, I tested both 4low and all. Both still says 3584 under GPU-z. I wonder...
yeah noticed that too... somewhere (anandtech) I read a reply from AMD marketing rep, that Fury Air always symmetrically disabled. what I see here suggested he was wrong.nonetheless, I think the day of R9 290 unlockables seems to be behind us, AMD seems intent on HW locks now (looking at the 285s), guessing this is mainly die salvaging from the MAC production line.
here's my FURY AIR (sapphire 100379SR), looks like hardware locked. Adapters detected: 1 Card #1 PCI ID: 1002:7300 - 174B:E329 00 Memory config: 0x00000000 Fiji-class chip with 16 compute units per Shader Engine SE1 hw/sw: 00030000 / 00000000 [..............xx] SE2 hw/sw: 00030000 / 00000000 [..............xx] SE3 hw/sw: 00810000 / 00000000 [........x......x] SE4 hw/sw: 00030000 / 00000000 [..............xx] 56 of 64 CUs are active. HW locks: 8 / SW locks: 0. Sorry, all 8...
A long time lurker, noob poster here. I been tinkering with PC since the i486 days, generally been quiet unless I have something to contribute. Finally got my R9 FURY Tri-X (sapphire non OC version) yesterday. I had some time benching and generally able to verify the online benches from Anand/Toms/TechReport etc. The next step was to test fury X bios, I grabbed & flashed both bios (AMD & ASUS) from...
New Posts  All Forums: