You mean software? Same as I have been with much better luck before updates.
Yes, I meant software, sorry.
Are you talking from a miner's perspective or a Pool mantainer perspective?
In any case the solution is the same.
Normally a miner goes to the getting Ready section of the pool, has a look at the configs and addresses the port with the lowest difficulty. After which he/she calls it a day.
If the software on pool's side works, and talks appropriately with mining software it may work, but this introduces an overhead to pool's brain that finds itself continuously calculating the appropriate diff for each connected miner, getting meanwhile a certain amount of shares which are simply Beyond its own calculating power.
A possible solution exists: educating the miners to set their own diff, according to their own hardware power.
This is done by carefully running their mining software distinguishing each pc of the rig from the others by the username.rig config.
But most important thing is to set a multiplier factor if using mining software that allows for it.
I.e., under ccminer-2.0-rc2-cuda-8.0 which supports Aeon, there is the
-m option that defines the multipling factor of the diff as agreed by handshake with pool.
A GTX1070 will have nearly the double of Aeon hashing power of a GTX1050Ti which in turn will be roughly 6/4 of a GTX960M. GTX940M will be 1/8th of GTX1070.
So if mining software tells you the handshake had been at Diff 1 for GTX940M, it will be a good idea to set it at
-m 1 and GTX1070 will be
-m 8 . As a starting point. The other ones will be set accordingly.
The connection will be done with the lowest starting diff port of the pool, which will be relieved from a lot of useless work.
Miner will see a much more constant hash rate (and more money).
