Topic: gpu error, i need help (Read 175 times)
install crimson 16.2 driver and select compute instead of gaming on the setup that will work. And dont forget to restart windows after crimson 16.2 installed
hi, my english is not that good but i can't find an answer for my issue with my rx580s on my language forum.
I have a particular problem with some rx580 4gb and 8gb (my first rig). The first time i built it, it has working at 24,1MH/s on ETH, then i tried OC but i only got 24,7MH/s, so i looked for a way to get the 30Mh/s that everyone had. I found information about BIOS mod and i did it, i copied the mem strap (timing)from 1500 to the 1625-1750-2000 fields on the 4gb and i tried to apply same method to the 8gb (by copying the 1/2:1750 to their respective fields) but it didn't work, so i used a third party mod. With the bios mod + OC i finally got the 30Mh/s.
I spent several days testing some oc values with -84mv (undervolt) to get the best config, but i had incompatibilities with the mobo (asus b250f strix) on the main pciex16 (when i used the slot, one gpu got an error when smos try to apply oc), but with windows 10 pro i solved it.
I didn't use a patched version of Crimson because i could oc the gpus without problem, but when the crimson beta for blockchain released, i installed it and i got an 0.5~1.5Mh/s aditional for each gpu. It worked great for 1 day, but the next day it crashed. I tried again and crashed again after 16hs, i keep trying changing some values of oc but every time it crashed faster.
Now i can't make the rig work with all the gpus connected at the same time, it loads windows perfectly and can play youtube normaly, but when it start to mining, a few gpus crashes and afterburner shows crazy numbers on the temperature and voltage on the crashed ones, then claymore's watchdog restart the miner, it freezes the console, and when i close it, it freezes Windows.
I checked all the gpus and tested one at time, 2/6 works fine but the other just crashed few minutes after start mining. I put the original bios back to see if that solves it, but nothing changed, i buyed new risers and mobo (asrock pro btc) to test if the gpus aren't the problem, but the same happens on windows and smos, with no oc/uv and origina bios.
I don't know what to do, i tried everything. There is still 2 month of warranty, they might fix the gpus but i don't thing it's a hw problem, they never get above 55ºC, never touched the power limit, and i only applied undervolt and 1300/2150 values of oc (2250 on the 8gb)
Can anyone help me? (sorry for my "engrish", if you cant understand some of this, just ask me about that part, or if you understand spanish it would be easier)
I have a particular problem with some rx580 4gb and 8gb (my first rig). The first time i built it, it has working at 24,1MH/s on ETH, then i tried OC but i only got 24,7MH/s, so i looked for a way to get the 30Mh/s that everyone had. I found information about BIOS mod and i did it, i copied the mem strap (timing)from 1500 to the 1625-1750-2000 fields on the 4gb and i tried to apply same method to the 8gb (by copying the 1/2:1750 to their respective fields) but it didn't work, so i used a third party mod. With the bios mod + OC i finally got the 30Mh/s.
I spent several days testing some oc values with -84mv (undervolt) to get the best config, but i had incompatibilities with the mobo (asus b250f strix) on the main pciex16 (when i used the slot, one gpu got an error when smos try to apply oc), but with windows 10 pro i solved it.
I didn't use a patched version of Crimson because i could oc the gpus without problem, but when the crimson beta for blockchain released, i installed it and i got an 0.5~1.5Mh/s aditional for each gpu. It worked great for 1 day, but the next day it crashed. I tried again and crashed again after 16hs, i keep trying changing some values of oc but every time it crashed faster.
Now i can't make the rig work with all the gpus connected at the same time, it loads windows perfectly and can play youtube normaly, but when it start to mining, a few gpus crashes and afterburner shows crazy numbers on the temperature and voltage on the crashed ones, then claymore's watchdog restart the miner, it freezes the console, and when i close it, it freezes Windows.
I checked all the gpus and tested one at time, 2/6 works fine but the other just crashed few minutes after start mining. I put the original bios back to see if that solves it, but nothing changed, i buyed new risers and mobo (asrock pro btc) to test if the gpus aren't the problem, but the same happens on windows and smos, with no oc/uv and origina bios.
I don't know what to do, i tried everything. There is still 2 month of warranty, they might fix the gpus but i don't thing it's a hw problem, they never get above 55ºC, never touched the power limit, and i only applied undervolt and 1300/2150 values of oc (2250 on the 8gb)
Can anyone help me? (sorry for my "engrish", if you cant understand some of this, just ask me about that part, or if you understand spanish it would be easier)
Jump to: