One more question plz.
The pool now implement VAR DIFF automatically. While can one fix a DIFF for each of his worker, manually and seperately?(Like some BTC pools I've mined with.)
Since I got a new problem: Due to unstable network, some share lost -> DIFF goes lower -> network load heavier --> more share lost --> DIFF even lower.
I think for some guys with a unstable network(like me), income variability is not so important as network load.
Usually when people ask for the ability to manually set share difficulty it's because they don't understand the underlying math, and naively think it's intrinsically better to have a low or high share difficulty.
I think this is the first time I hear a semi-valid justification for such a feature. However, if your network is
that unstable, then you probably have much more serious problems (such as the amount of time you spend hashing stale data) than those caused by an unsuitably low share difficulty.
what is the average stale rate for you guys? my gpus (7950s') are all averaging around 1.2-1.3% stale rate and was wondering if it is a little high?
The pool average right now is 0.5%, with some users going as low as 0.03%.
However, software latency is sometimes unavoidable when GPU mining, so a stale rate around 1% is usually considered acceptable for GPU systems. Please note that this kind of latency is not due to your network connection or to the pool, but to the miner itself.