Welcome to the Dash Forum!

Please sign up to discuss the most innovative cryptocurrency!

sgminer v5 troubles

Discussion in 'Mining' started by 5kmi, Sep 9, 2014.

  1. 5kmi

    5kmi Member

    Joined:
    Jul 21, 2014
    Messages:
    41
    Likes Received:
    19
    Trophy Points:
    58
    Background info: win7 64bit, amd 13.12 drivers with 14.6 ocl files dropped into sgminer folders for bin creation. 7790 x11 hashrate: 1.7-1.75mh, 270 x11 hashrate: 2.6-2.75mh

    So I've been testing sgminer v5 (nicehash release) on my PC for the past week or so, using multialgo switching between x11, x13, x15 and nist5, on a single MSI 270 to great success. The trouble started a couple days ago when I tried to do the same on a couple of my multigpu rigs. Tried x11 on my "spare parts rig" which consists of a msi270 and a msi7790, and the results were odd... 7790 producing about 20% of expected hashrate and 270 making about 60% of expected hashrate, today I discovered turning off the 7790 made the 270 go up to full speed, while turning off 270 yielded no change to 7790 hashrate. CPU and ram are not bottlenecking as far as I can tell, this rig had 2gb ddr3 and I swapped it out for a 4gb stick to no difference. Tried on another rig with 4x 7790 and 1 270 (Asus 970mobo, fx-6300 processor, 4gb ram) and got even stranger results. 7790s were producing 5-30% of expected hashrate, each card producing a different value, and 270 making about 50% of expected value. After a few minutes of this I got a bsod, which afterwards occured anytime i tried to mine any algo, with any software, with any gpu in the system until wiped amd drivers, cleaned registry, and reinstalled (whole other story there). Even with fresh drivers I still get the same results with sgminer v5 as before, but thankfully no bsod. Both machines are back to p2pool mining darkcoin with the sgminer_14.6 release, with amdocl files in folder for bin creation.

    If anyone has had similar issues with this release I have yet to find any other posts about it, and I still find it odd that v5 runs perfectly on my single GPU PC and when I have only 1 270 enabled on multigpu machines. Any ideas?

    Conf file of the spare parts rig:
    {
    "pools" : [

    {
    "name" : "X11_multi",
    "url" : "stratum+tcp://stratum.westhash.com:3336",
    "user" : "mybtcaddress",
    "pass" : "f0=0;f2=0;f3=6.35;f4=4.75;f5=0;f6=4.2;f7=20",
    "nfactor" : "10",
    "algorithm" : "darkcoin-mod"
    },
    {
    "name" : "X13_multi",
    "url" : "stratum+tcp://stratum.westhash.com:3337",
    "user" : "mybtcaddress",
    "pass" : "f0=0;f2=0;f3=6.35;f4=4.75;f5=0;f6=4.2;f7=20",
    "nfactor" : "10",
    "algorithm" : "marucoin-mod"
    },
    {
    "name" : "X15_multi",
    "url" : "stratum+tcp://stratum.westhash.com:3339",
    "user" : "mybtcaddress",
    "pass" : "f0=0;f2=0;f3=6.35;f4=4.75;f5=0;f6=4.2;f7=20",
    "nfactor" : "10",
    "algorithm" : "bitblock"
    },
    {
    "name" : "NIST5_multi",
    "url" : "stratum+tcp://stratum.westhash.com:3340",
    "user" : "mybtcaddress",
    "pass" : "f0=0;f2=0;f3=6.35;f4=4.75;f5=0;f6=4.2;f7=20",
    "nfactor" : "10",
    "algorithm" : "talkcoin-mod"
    },
    {
    "url" : "stratum+tcp://54.186.8.140:7903",
    "user" : "drkaddress",
    "pass" : "x",
    "nfactor" : "10",
    "algorithm" : "darkcoin-mod"
    },
    {
    "url" : "stratum+tcp://darkcoin.kicks-ass.net:7903",
    "user" : "drkaddress",
    "pass" : "x",
    "nfactor" : "10",
    "algorithm" : "darkcoin-mod"
    },
    {
    "url" : "stratum+tcp://drk.coinmine.pl:16090",
    "user" : "user.worker",
    "pass" : "x",
    "nfactor" : "10",
    "algorithm" : "darkcoin-mod"
    }

    ],
    "device" : "0,1",
    "intensity" : "17",
    "worksize" : "64",
    "lookup-gap" : "2",
    "thread-concurrency" : "8192",
    "gpu-threads" : "2",
    "gpu-engine" : "900-1130",
    "auto-gpu" : true,
    "gpu-fan" : "30-75",
    "auto-fan" : true,
    "gpu-memclock" : "1200,1400",
    "gpu-memdiff" : "0,0",
    "gpu-powertune" : "0",
    "gpu-vddc" : "0.000",
    "temp-cutoff" : "85",
    "temp-overheat" : "82",
    "temp-target" : "70",
    "api-mcast-port" : "4028",
    "api-port" : "4028",
    "expiry" : "5",
    "failover-only" : true,
    "failover-switch-delay" : "60",
    "gpu-dyninterval" : "1",
    "gpu-platform" : "0",
    "log" : "5",
    "no-pool-disable" : true,
    "queue" : "0",
    "scan-time" : "5",
    "tcp-keepalive" : "30",
    "temp-hysteresis" : "3",
    "shares" : "0"
    }

    Just tried compiling bins without the amdocl files to see if that was involved with the problem, hashrate slightly lower no both cards, but both still not producing anywhere near what they should, unless I turn the 7790 off.
     
  2. fra55

    fra55 Member

    Joined:
    Mar 26, 2014
    Messages:
    143
    Likes Received:
    14
    Trophy Points:
    68
    I have a similar problem with my R9 280x and 6850. With the R9 working, the 6850 produces about 80% of expected shares, but it's acceptable.
    I still haven't figured out the reason, maybe a driver problem when using cards with different chips