I too have had to take my November Jupiter to another pool. I got over 10% hardware errors and cores enabling/disabling constantly. After my 12hr avg @ Eligius dropped below 600, I switched to BTC Guild and now get 660-690 at the pool with 1.7% hardware errors and no cores goin on and off.
I noticed that my diff @ Eligius alternated between 512 and 1024 while I consistently get 256 @ BTC Guild, if that might have anything to do with it.
I noticed that my diff @ Eligius alternated between 512 and 1024 while I consistently get 256 @ BTC Guild, if that might have anything to do with it.
Yep same here. Oct jupiter works flawlessly on eligius but Nov jupiter has 8% hw errors and cores disabling like crazy but when I switch Nov jupiter to ghash pool it works fine errors drop back and no disabled cores
My November Jupiter is working fine (660 GH/s with 2% HW errors and 0.1% rejects), although I am using the cgminer-tune binary from here and I am using --load-balance to split my hashing between btcguild and eligius.
My only problem is that when the miner initially starts up, there are a few moments of insanity until the first vardiff adjustment makes the share difficulty reasonable. I wish there was a way to specify a starting difficulty other than '2'. I'd probably pick something like 64. Then my Jupiter and BitFury rigs would vardiff up to something like 256/512 and my BFL Singles would vardiff down to 16/32 as appropriate. I gather that wizkid057 is working on an enhancement to bfgminer, which is fine, but won't help me on devices that have their own mining software (e.g. my BitFury).
