I benchmarked some GPUs in my rigs with Genoil ethminer 1.1.7 in range of epochs and got interesting results.
Tested with DAG in blocks: 1K, 2M, 3M, 4M
1. 270x 1000 core/1400mhz ram
2. 280x 970/1600mhz
3. 290x 900/1000mhz
Pitcairn (270x, 7870) will stop mining in the range of blocks 2.8-2.9M
Thaiti (280x, 7970) will become inefficient in the near future. Where is the bottleneck?
Hawaii (290..390x) shows very stable hashrate
Interesting to know results of other graphic cards, and which of them are not susceptible to degradation of hashrate. Perhaps they are R9 380, RX 470-480, R9 Fury and Nano.
Run ethminer in Benchmarking mode:
x: Select which OpenCL devices to mine on
n: specify block number to benchmark against specific DAG.
Tested with DAG in blocks: 1K, 2M, 3M, 4M
1. 270x 1000 core/1400mhz ram
2. 280x 970/1600mhz
3. 290x 900/1000mhz
| Block\GPU | 270x | 280x | 290x |
| 1K, very beginning | 18.1 mh/s | 25.1 mh/s | 24.4 mh/s |
| 2M, recently | 11.2 | 16.8 | 24.4 |
| 3M, the end of 2016 | 0 | 13.3 | 24.1 |
| 4M, about a year from now | 0 | 11.5 | 24.1 |
Pitcairn (270x, 7870) will stop mining in the range of blocks 2.8-2.9M
Thaiti (280x, 7970) will become inefficient in the near future. Where is the bottleneck?
Hawaii (290..390x) shows very stable hashrate
Interesting to know results of other graphic cards, and which of them are not susceptible to degradation of hashrate. Perhaps they are R9 380, RX 470-480, R9 Fury and Nano.
Run ethminer in Benchmarking mode:
ethminer -G --opencl-devices <0 1 ..x> -M <n>x: Select which OpenCL devices to mine on
n: specify block number to benchmark against specific DAG.