OPM
[pure testing] testing the drift compensation
will wait the benchmark are updated to trigger the benchmark. we are experiencing drift compensation worsening running of some aquifer related cases, so hoping to test it with other benchmark cases.
Alternatively, we can also have more sophisticate strategy to trigger the drift compensation.
Benchmark result overview:
| Test | Configuration | Relative |
|---|---|---|
| opm-git | OPM Benchmark: drogon - Threads: 1 | 0.99 |
| opm-git | OPM Benchmark: drogon - Threads: 8 | 0.788 |
| opm-git | OPM Benchmark: smeaheia - Threads: 1 | 1.042 |
| opm-git | OPM Benchmark: smeaheia - Threads: 8 | 1.029 |
| opm-git | OPM Benchmark: spe10_model_1 - Threads: 1 | 0.993 |
| opm-git | OPM Benchmark: spe10_model_1 - Threads: 8 | 1.1 |
| opm-git | OPM Benchmark: flow_mpi_extra - Threads: 1 | 0.999 |
| opm-git | OPM Benchmark: flow_mpi_extra - Threads: 8 | 1.012 |
| opm-git | OPM Benchmark: flow_mpi_norne - Threads: 1 | 0.998 |
| opm-git | OPM Benchmark: flow_mpi_norne - Threads: 8 | 0.979 |
- Speed-up = Total time master / Total time pull request. Above 1.0 is an improvement. *
FAILED RUNS: OPM Benchmark: punqs3 - Threads: 1 - 2023-06-07 Master Reference: The test run did not produce a result. E: Error: Unrecoverable errors while loading input: Problem with keyword OPM Benchmark: punqs3 - Threads: 1 - opm-simulators 4696: pure testing testing the drift compensation: The test run did not produce a result. E: Error: Unrecoverable errors while loading input: Problem with keyword OPM Benchmark: punqs3 - Threads: 8 - 2023-06-07 Master Reference: The test run did not produce a result. E: Error: Unrecoverable errors while loading input: Problem with keyword OPM Benchmark: punqs3 - Threads: 8 - opm-simulators 4696: pure testing testing the drift compensation: The test run did not produce a result. E: Error: Unrecoverable errors while loading input: Problem with keyword OPM Benchmark: flow_mpi_norne_4c_msw - Threads: 1 - 2023-06-07 Master Reference: The test run did not produce a result. E: Error: Unrecoverable errors while loading input: Problem with keyword OPM Benchmark: flow_mpi_norne_4c_msw - Threads: 1 - opm-simulators 4696: pure testing testing the drift compensation: The test run did not produce a result. E: Error: Unrecoverable errors while loading input: Problem with keyword OPM Benchmark: flow_mpi_norne_4c_msw - Threads: 8 - 2023-06-07 Master Reference: The test run did not produce a result. E: Error: Unrecoverable errors while loading input: Problem with keyword OPM Benchmark: flow_mpi_norne_4c_msw - Threads: 8 - opm-simulators 4696: pure testing testing the drift compensation: The test run did not produce a result. E: Error: Unrecoverable errors while loading input: Problem with keyword
View result details @ https://www.ytelses.com/opm/?page=result&id=2120
will wait the benchmark are updated to trigger the benchmark. we are experiencing drift compensation worsening running of some aquifer related cases, so hoping to test it with other benchmark cases.
I have looked a bit more on these cases and I am not sure if the cause of the worsening convergence is related to drift compensation itself. Note that to get a good comparison on the performance effect of the drift-compensation you should test the combination of tighter mb tolerance and drift compensation false with default mb tolerance and drift compensation true. I.e. this Pr should be updated with a change to the default mb tolerance set to 1e-7 to give a fear comparison.
Benchmark result overview:
| Test | Configuration | Relative |
|---|---|---|
| opm-git | OPM Benchmark: drogon - Threads: 1 | 1.012 |
| opm-git | OPM Benchmark: drogon - Threads: 8 | 0.604 |
| opm-git | OPM Benchmark: punqs3 - Threads: 1 | 0.974 |
| opm-git | OPM Benchmark: punqs3 - Threads: 8 | 1.104 |
| opm-git | OPM Benchmark: spe10_model_1 - Threads: 1 | 0.999 |
| opm-git | OPM Benchmark: spe10_model_1 - Threads: 8 | 1.099 |
| opm-git | OPM Benchmark: flow_mpi_extra - Threads: 1 | 1.01 |
| opm-git | OPM Benchmark: flow_mpi_extra - Threads: 8 | 0.849 |
| opm-git | OPM Benchmark: flow_mpi_norne - Threads: 1 | 0.991 |
| opm-git | OPM Benchmark: flow_mpi_norne - Threads: 8 | 1.011 |
| opm-git | OPM Benchmark: flow_mpi_norne_4c_msw - Threads: 1 | 0.837 |
| opm-git | OPM Benchmark: flow_mpi_norne_4c_msw - Threads: 8 | 1.041 |
- Speed-up = Total time master / Total time pull request. Above 1.0 is an improvement. *
FAILED RUNS: OPM Benchmark: smeaheia - Threads: 1 - opm-simulators 4696: pure testing testing the drift compensation: The test run did not produce a result. E: mpirun was unable to launch the specified application as it could not access OPM Benchmark: smeaheia - Threads: 8 - opm-simulators 4696: pure testing testing the drift compensation: The test run did not produce a result. E: mpirun was unable to launch the specified application as it could not access
View result details @ https://www.ytelses.com/opm/?page=result&id=2128