Forum Home
    • Register
    • Login
    • Search
    • Recent
    • Tags
    • Popular

    HD7970 vapor-x 6GB only 650KH stable

    Mining
    4
    9
    1210
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • M
      MindaugasC last edited by

      Hello guys,

      trying to put my 7970 vapor-x edition to more then 680KH/s speed, but unlucky.

      Right now, i have stable 650KH/s

      [code]“expiry” : “20”,
      “failover-only” : true,
      “fix-protocol” : true,
      “hotplug” : “5”,
      “kernel-path” : “/usr/local/bin”,
      “log” : “5”,
      “queue” : “0”,
      “scan-time” : “10”,
      “scrypt” : true,
      “shares” : “0”,
      “auto-fan” : true,
      “gpu-threads” : “1”,
      “gpu-engine” : “0-1231”,
      “gpu-fan” : “0-80”,
      “gpu-platform” : “0”,
      “gpu-memclock” : “1850”,
      “gpu-memdiff” : “0”,
      “gpu-powertune” : “20”,
      “gpu-vddc” : “0.000”,
      “intensity” : “13”,
      “temp-target” : “70”,
      “temp-overheat” : “80”,
      “temp-cutoff” : “85”,
      “temp-hysteresis” : “3”,
      “vectors” : “1”,
      “worksize” : “256”,
      “lookup-gap” : “2”,
      “shaders” : “2048”,
      “thread-concurrency” : “8192”,
      “no-pool-disable” : true[/code]

      If i change TC to higher when 8192 - KH speed is decreasing 450-550KH
      If i change gpu-threads to 2 - KH speed is decreasing to 390KH
      gpu-engine - lower - decreasing. Higher - system crashing.
      memclock 1850 working stable, with higher - system crashing.

      I have only one idea why im getting only 650KH stable, my motherboard is very old (using an old unused PC)
      [url=http://www.gigabyte.com/products/product-page.aspx?pid=2810#ov]http://www.gigabyte.com/products/product-page.aspx?pid=2810#ov[/url]
      GA-X48-DS4

      Any ideas?

      [attachment deleted by admin]

      1 Reply Last reply Reply Quote 0
      • A
        another_user last edited by

        you need to find why you have hardware errors. something is not working right, and will for sure cause problems.

        1 Reply Last reply Reply Quote 0
        • M
          MindaugasC last edited by

          [quote name=“another_user” post=“39970” timestamp=“1386236384”]
          you need to find why you have hardware errors. something is not working right, and will for sure cause problems.
          [/quote]

          i tryed everything…TC, gpu-engine, memclock and so on. This is my best config…

          1 Reply Last reply Reply Quote 0
          • A
            another_user last edited by

            you mention you are using an old pc. is the power supply up to the task?

            1 Reply Last reply Reply Quote 0
            • M
              MindaugasC last edited by

              [quote name=“another_user” post=“40011” timestamp=“1386241158”]
              you mention you are using an old pc. is the power supply up to the task?
              [/quote]

              Yes,

              1100W

              1 Reply Last reply Reply Quote 0
              • R
                RIPPEDDRAGON Regular Member last edited by

                GPU mem clock is WAY TOO HIGH at 1850, it looks like you are letting the gpu core clock throttle itself as well. Read the cgminer guide and follow it exactly and these problems wont happen.

                Make sure you have these in your .bat file before the cgminer config, this will allow for higher a TC.
                setx GPU_MAX_ALLOC_PERCENT 100
                setx GPU_USE_SYNC_OBJECTS 1

                Then run it without setting a TC and see what cgminer sets. 7970s want to run at I13 somewhere between 1040-1130 gpu core clock and around 1500 memory clock. Its not about smashing forward pushing an OC to its max, its about finding the right BALANCE between core and mem speeds and finding the best TC for your card. With this much vram you may be able to run a TC number higher than 24000.

                1 Reply Last reply Reply Quote 0
                • M
                  MindaugasC last edited by

                  [quote name=“RIPPEDDRAGON” post=“40080” timestamp=“1386251987”]
                  GPU mem clock is WAY TOO HIGH at 1850, it looks like you are letting the gpu core clock throttle itself as well. Read the cgminer guide and follow it exactly and these problems wont happen.

                  Make sure you have these in your .bat file before the cgminer config, this will allow for higher a TC.
                  setx GPU_MAX_ALLOC_PERCENT 100
                  setx GPU_USE_SYNC_OBJECTS 1

                  Then run it without setting a TC and see what cgminer sets. 7970s want to run at I13 somewhere between 1040-1130 gpu core clock and around 1500 memory clock. Its not about smashing forward pushing an OC to its max, its about finding the right BALANCE between core and mem speeds and finding the best TC for your card. With this much vram you may be able to run a TC number higher than 24000.
                  [/quote]

                  Always doing it:
                  setx GPU_MAX_ALLOC_PERCENT 100
                  setx GPU_USE_SYNC_OBJECTS 1

                  I read SCRYPT-README file it saying, first need to get memory to higher clock as possible and later core-clock, and only after it - TC and I.

                  I tryed all possible gpu core clock with mem clock of 1500. max hash was 390KH/s

                  1 Reply Last reply Reply Quote 0
                  • T
                    TheTribesman last edited by

                    I can see why you want to increase hashrate but I personally feel that stability is just as important, if not more important, than just pure hashrate. This is for purely practical reasons… I have to leave my house sometimes!

                    My rig has 3x7950 cards that can get up to 702KHz each (it’s in my attic/loft, I use XUbuntu/cgminer - monitored remotely). I tuned them up to that so I was getting a combined 2.04Mhz. That was great except that they kept crashing every few hours, or 1 card might stop, loads of invalids, or something else might cause the rig to stop. I was constantly monitoring the rig. I had no confidence that I’d get back from work and it’d still be running. For what? For a few extra coins/day? I wasn’t even achieving that goal with a rig that was only running only about half the day. I was constantly tuning, and retuning, and watching etc…

                    In the end, I downgraded them to run at 650-660KHz each, and now I can leave my rig running 24/7 and not worry about it.

                    “intensity” : “20,20,20”,
                    “vectors” : “1,1,1”,
                    “worksize” : “256,256,256”,
                    “kernel” : “scrypt,scrypt,scrypt”,
                    “lookup-gap” : “2,2,2”,
                    “thread-concurrency” : “40960,40960,40960”,
                    “shaders” : “0,0,0”,
                    “gpu-engine” : “1150,1150,1150”,
                    “gpu-memclock” : “1500,1500,1500”,
                    “gpu-powertune” : “20,20,20”,
                    “temp-target” : “80,80,80”,
                    “log” : “5”,
                    “no-pool-disable” : true,
                    “queue” : “1”,
                    “scan-time” : “60”,
                    “scrypt” : true,
                    “shares” : “0”

                    1 Reply Last reply Reply Quote 0
                    • M
                      MindaugasC last edited by

                      [quote name=“TheTribesman” post=“40115” timestamp=“1386257408”]
                      I can see why you want to increase hashrate but I personally feel that stability is just as important, if not more important, than just pure hashrate. This is for purely practical reasons… I have to leave my house sometimes!

                      My rig has 3x7950 cards that can get up to 702KHz each (it’s in my attic/loft, I use XUbuntu/cgminer - monitored remotely). I tuned them up to that so I was getting a combined 2.04Mhz. That was great except that they kept crashing every few hours, or 1 card might stop, loads of invalids, or something else might cause the rig to stop. I was constantly monitoring the rig. I had no confidence that I’d get back from work and it’d still be running. For what? For a few extra coins/day? I wasn’t even achieving that goal with a rig that was only running only about half the day. I was constantly tuning, and retuning, and watching etc…

                      In the end, I downgraded them to run at 650-660KHz each, and now I can leave my rig running 24/7 and not worry about it.

                      “intensity” : “20,20,20”,
                      “vectors” : “1,1,1”,
                      “worksize” : “256,256,256”,
                      “kernel” : “scrypt,scrypt,scrypt”,
                      “lookup-gap” : “2,2,2”,
                      “thread-concurrency” : “40960,40960,40960”,
                      “shaders” : “0,0,0”,
                      “gpu-engine” : “1150,1150,1150”,
                      “gpu-memclock” : “1500,1500,1500”,
                      “gpu-powertune” : “20,20,20”,
                      “temp-target” : “80,80,80”,
                      “log” : “5”,
                      “no-pool-disable” : true,
                      “queue” : “1”,
                      “scan-time” : “60”,
                      “scrypt” : true,
                      “shares” : “0”
                      [/quote]

                      I’m understanding, stability in a first place. My configuration which im getting 650KH is stable, but with HW (~10 per 24h).
                      Also im understanding, this kind of card, can get really better KH speed…

                      I can give teamviewer login/pass.

                      Paying 50FTC who make my card work stable at 700KH
                      Paying 75FTC who make my card work stable at 720KH

                      1 Reply Last reply Reply Quote 0
                      • First post
                        Last post