Hi, ethermine is reporting an average of 125 MH/s for my 5x Sapphire Nitro RX 470 4gb rig. The panel of ethosdistro is reporting 137 MH/s for this rig, maybe there are some configuration wrong? I readed that a correct value for farm-recheck increase the average hashrate and reduces stale shares. I have this values in config file:
flags --cl-global-work 8192 --farm-recheck 200
I don't have a value for local work, how can i determinate properly the value for farm-recheck? Thanks in advance.
PS: i found this formula for determinate kernel time but i have no value for local work
1000/your_GPU_hash_rate, example: 1000/15 = 66.67 - that's the number of nanoseconds it takes you to do 1 individual hash.
Multiply that times global_work times local_work: 66.67 * 16384 * 64 = 69905067 ns per kernel
Divide by 1M (move decimal point 6 digits to the left) to get milliseconds= 69.9 ms kernel run time.
Thanks in advance.
flags --cl-global-work 8192 --farm-recheck 200
I don't have a value for local work, how can i determinate properly the value for farm-recheck? Thanks in advance.
PS: i found this formula for determinate kernel time but i have no value for local work
1000/your_GPU_hash_rate, example: 1000/15 = 66.67 - that's the number of nanoseconds it takes you to do 1 individual hash.
Multiply that times global_work times local_work: 66.67 * 16384 * 64 = 69905067 ns per kernel
Divide by 1M (move decimal point 6 digits to the left) to get milliseconds= 69.9 ms kernel run time.
Thanks in advance.