I'm an ex network admin with a MCSE, so not a complete cabbage!
No problem bro, i cannot know if people who ask questions here are very newbie or world-class experts
Miner show that card have 4096Mb memory, but can't use it all...
In GPU-Z you can get the maximum that can be allocated by OpenCL, and it's not 100% of VRAM. The environment settings at 100 help but not completely. Even on V7, JCE needs more memory than Claymore did.
GPU_FORCE_64BIT_PTR 100
this should be 0 or 1, and I advise not to set that variable at all, the drivers will choose the best one.
in new version somehow it detect iGPU as GPU 0, so try to change your config from GPU 1.
That's a
test mode introduced by the 0.33a where I unlocked all GPUs. I will lock back all but AMD and Intel, since mining on Intel works (maybe not fast, but it mines).
Abort was called at 318 line in file:
D:\qb\workspace\19992\src\vpg-compute-neo\runtime\os_interface\windows\wddm.cpp
Press any key to continue . . .
That looks like a crash in a driver, but I never reproduced it. Does it happen at mining start?
Hey. RX480 video card 8GB nitro + error "Compilation of OpenCL kernels failed.
what version of drivers and Windows?
Why in newer versions the hashrate 300-580 jumps
add parameter --no-warmup
heavy should perform just like the old version
version 0.33b2 adds a very little CN-Heavy optim, unrelated to Monero v8
Tested on 5*rx550 baffin unlocked monero v8 (actually v9): 330 h/s each on "--auto"
SRB gives 430 h/s = the same as monero v7.
Adrenalin 18.6.1
Win10x64
FX8320
Wanna try not auto mode, but idk what's that:
"mode" : "GPU", -ok, gpu...
"worksize" : 8,
"alpha" : 64,
"beta" : 8,
"gamma" : 4,
"delta" : 4,
"epsilon" : 4,
"zeta":4,
"index" : 0, - number of gpu
"multi_hash":128
the multi_hash on a 2G card should be rather 400 or more, on two threads. And 430 for v7 is low, whatever miner you use.
The --auto in GPU mode is made to be safe, not fast. To get max speed, use the manual config.
The CPU autoconfig aims for speed. That's because you hardly can crash the rig by mining on CPU, but overloading the GPU can be dangerous, even for your PSU.
greeks above beta are obsolete. worksize is like in xmrstak. multi_hash is number of GPU threads, each thread needs 2M of VRAM (or 4M for heavy, or 1M for Light). alpha and beta change some inner parallelization settings.
