Hi Guys! Like the project, but I've got a bit of an odd issue I'm hoping you can help with.
On my multi-card rig, some sort of error occurred during benchmarking. This has resulted in the UI perpetually showing "Benchmarking - Target: (value)/day (ETN+PIRL)" despite the fact that it's not benchmarking anything. Instead, it completely benchmarked the one Nvidia card, but all 4 AMD cards have been on ZEN and ZEN only since I started. Checking the Benchmark history, it just didn't even try to bench anything else for the AMD cards.
Interestingly, I tried uninstalling and reinstalling, thinking this would force it to re-benchmark everything, but it seems that once again the AMD cards are straight to mining zen while the Nvidia goes through a series of benchmarks.
Am I nuts? On my other rig, a single AMD card, it benchmarked as expected and has switched here and there as profits moved around.
Is there any way to ensure I have truly knocked out all previous benchmarks and really start from a clean slate on the multi-card rig?
Welcome aboard!
Here is a quick explanation about benchmarking on WinMiner -
(1) Benchmarking - Target - This metric shows up before benchmarking takes place, and it is based on crowd wisdom logic - it analyzes the performance of very similar hardware by other users on WinMiner and gives out an expected target metric. The idea is that users will have an estimate of what realistically they can expect before the full cycle of benchmarking is completed.
(2) Benchmark modes -
(a) Express (default) - in this mode the benchmarking will commence with the combination most likely to provide the best performance for the specific hardware, and then once in a while attempt a different combination to make sure the best combination is indeed the one being mined. This mode gets a user up and running with max performance asap.
(b) Regular Benchmark - in this mode all combinations will be tried out, one by one until the most profitable combination is chosen
In addition, once every 24 hours hardware will be examined per each algo for hash poer performance.
Hope this clarifies :-)
