Quote from: Freight on February 16, 2018, 07:44:18 AM

I do not think that I have to wait every time 5 times more time to know each time that the order is the same and only the best hashrate is saved, but indeed I really think it would be useful to have an option like "Test all miners and pick the fastest miner for each algo", right ?

Now, the question remains what do we do when we benchmark today and we have the values saved and when we benchmark again 2 weeks later the values are different. The values also lead to wrong algos ordering because new hashrates changed the ordering, making some algos more profitable.

It is also useful to have the latest benchmark date saved and schedule periodic benchmarks. I noticed that benchmark hash rates varies for the same RIG when benchmarking again after 1 week.

Any ideea ?

I also would be interested in automated benchmarking every so often to guarantee the best profits.
[/quote]

Is Patrike alive ? It seems other software will soon be better than Awesome Miner if AM is not maintained and improved every day.

I strongly advise that AM should have an option for "precise benchmark" that should also test various intensity settings for each GPU and each mining software and save them. Right now AM leaves intensity settings to default miner's values that results in 20% lower income from the miners just because of lacking the usage of the intensity with the best results.

What news do we have about better results with pool statistics to avoid issues with spikes ?

BTW, did you see the features other software offers that is missing from AwesomeMiner ?
https://github.com/nemosminer/NemosMiner-v2.5

NemosMiner
Uses calculations based on 24hractual and currentestimate ahashpool prices to get more realistic estimate.
  Includes some trust index based on past 1hr currentestimate variation from 24hr.
  AND is NOT sensible to spikes.
  This shows less switching than following Current Estimate and more switching that following the 24hr Actual.
  Better profitability.