Quote from: fellowbeing on July 23, 2018, 01:26:43 PM

Quote from: doktor83 on July 22, 2018, 06:05:31 AM

Quote from: fellowbeing on July 22, 2018, 02:02:41 AM

Vega FEs are allergic to 18.3.4 and I'll get less hash-rate on Alloy/Heavy and CN7, on Alloy there is alot of STALE shares too, so the only viable driver is 18.6.1.

Results for a rig having 5 Vega FEs SRB Miner 1.6.4:
driver 18.6.1
CN7 = 11060 H/s
Alloy = 5670 H/s
note: Mining Alloy, on DevFee we get miner crash or error

driver 18.3.4
CN7 = 10900 H/s
Alloy = 5580 H/s
note: Mining Alloy, there is a lot of STALE shares.

Results for a rig having 5 Vega FEs SRB Miner 1.6.1:
driver 18.6.1
CN7 = 10910 H/s
Alloy = 5630 H/s

driver Blockchain Aug 23
CN7 = 10910 H/s
Alloy = 5630 H/s

=========================================
So in conclusion:
18.3.4 won't work for Vega FEs. For Alloy we have to use 1.6.1 and we get similar hashrates as blockchain driver.

Edited: The said crash/error occurs not on DevFee but right after when resuming!

Stales are not a problem if they are accepted,i guess the difficulty on the pool you use is low, and the pool sends jobs quickly, so that is why you get a lot of stales, but if they are accepted then you have no problem.
Also could you turn on logging (use --logfile log.txt in bat added) maybe i can get more info on the crash ?
If i understand you correctly only on 18.6.1 is the miner crashing, on 18.3.4 it isnt ?

Yes, I'll do this and see what we get. Regarding the drivers, I think 18.3.4 is just not working for me, even on CN7 I get lower hash-rate simiar to what I used to get on Blockchain driver about 150 H/s less or 50 H/s less on heavy/Alloy, but 18.6.1 has been very stable for me. So I avoid 18.3.4 totally. As for heavy/Alloy algo while using the SRB 1.6.4, I get GPU crash right after DevFee sometimes recovers after resetting the miner but other times I have to close the miner and restart it manually. I noticed this happened on 1.6.1 too but apparently less often. I don't have a clue what the cause could be, but I guess it might be something to do with change of algo from Alloy to DevFee and back. Unfortunately I have not had the chance to try any other heavy algo yet. I'll set the log and see what happens and come back. Thanks

There is no change of algo when doing devfee mining, if you mine alloy then the devfee is also doing alloy, if mining heavy then its heavy etc.
Also did you try switching pools manually (by defining more than one pool in pools.txt and pressing p) ? Is this crashing the miner ?