Re: [ANN] sph-sgminer: multi-coin multi-algorithm GPU miner

Quote from: DJstone on March 28, 2014, 02:39:35 PM

Quote from: ntsdm on March 28, 2014, 01:22:15 PM

is any one running a 7790 for quark? I cant get it ro run? Huh

Yes, I'm running 2x7970!!

Quote from: boubou on March 28, 2014, 12:31:56 PM

Quote from: DJstone on March 28, 2014, 12:07:30 PM

Quote from: CryptoGretzky on March 28, 2014, 10:08:57 AM

Quote from: DJstone on March 28, 2014, 07:12:01 AM

Hello,

I want to know, if you tested badmans and my version of SGminer already.
Pls let me (us) know

I have tested the latest badmans version.

I just tested your version. It did the same thing. It dropped after a while. I don't know if it has to do with the difficulty change of the block. That at least was the thing I saw when the rate started dropping.

Pls can you give further information to me?

What Pool?
What Configurations?
What Dsiplaydriver version are you using?
Can u check (maybe over Afterburner) if the Coreclock or the Memclock are changing while the card reduces by 10%?
Would u please make a log file till the reduction of speed? (make a start.bat pointing towards your sgminer.exe like this : "sgminer -k bla -u bla -p bla -o bla://bla.bla 2> log.txt")

Core clock and memclock are stable, only the hasrate drops
Bug happened on all pools
Bug happened with all configurations (even with no overclock)
Bug happened with catalyst 1.13 and 1.12 on linux bamt (other have the same bug on w7)

What is your version?
I'll try to male a log file.

I'm pretty confident there is a bug somewhere.
Question is: what can possibly make the hasrate to drop if the engine and memclock did not change?
If we have a list, may be we can track down the various possibility, but we have no list. I was thinking that only the engone and memory clock can make the hasrate moving (may be intensity too).

Is it possible that the intensity switch, and How can I track that?
Is there some other things that can modify the hasrate besides engine, memclock, intensity?

Ok, I really cannot beleave, that there is a bug in the code, that only appears on the 290.
If there would be any bug, related to that R9 290 behaviour, than it would be all cards (imho).

So maybe it has to do with the AMD driver, or with the BIOS on the Cards.

Would you please make a log file with verbose mode?
You have to add an:
"verbose" : true,
to your config or --verbose to your command line!

When I talk abot a bug, it's the fact that there is a probel that is only triggered by 290.
Bios cold be an issue, but among the 5 of us that encounter that bug (each one with several cards) we have various brand of 290 and various OS...

I'll do the test with the verbose option.