CUDA miner

1495052545569

Comments

  • dlehenkydlehenky Member Posts: 2,249 ✭✭✭✭
    @Marvell9 What I like about the Nanos is, come summer, with the undervolt bios already in place, I can underclock them all the way to 80-90 watts, which I know you already know. That's a lot of flexibility, which I may need when the heat comes. I don't know that much about the 390s; can you do that with them, too?
  • dlehenkydlehenky Member Posts: 2,249 ✭✭✭✭
    @wirelessnet2 The bios for the Nanos also undervolts memory -100mv, which you can't do with software. You can't do any undervolting on linux, which is what I run, so flashing bios is the answer, and a very good one, I might add.
  • workwork Member Posts: 2,084 ✭✭✭✭
    $390 for used nanos is pretty good. $200 for a refurb R9 290 sounds pretty good too tho... can always downclock and undervolt them too.
  • wirelessnet2wirelessnet2 Member Posts: 385 ✭✭✭
    edited May 2016
    @dlehenky how TF r u getting nanos for $390
  • dlehenkydlehenky Member Posts: 2,249 ✭✭✭✭
  • Marvell9Marvell9 Member Posts: 593 ✭✭✭
    @wirelessnet2 Used/refurbished from my supplier some from eBay
  • wirelessnet2wirelessnet2 Member Posts: 385 ✭✭✭
  • wirelessnet2wirelessnet2 Member Posts: 385 ✭✭✭
    @Marvell9 how do people get "suppliers"? kinda like...
  • Marvell9Marvell9 Member Posts: 593 ✭✭✭
    @wirelessnet2 when u buy 50k worth of hardware ..
  • chrizzochrizzo GermanyMember Posts: 131 ✭✭
    anyone some idea? My 390X get only 28M/h, but the most have 30-35 M/h? Using Win7 x64, crimson 15.12.
  • deliciousminingdeliciousmining Member Posts: 49
    Guys, are all Fiji cores so flexible? Can you make Fury X / Fury into Nano if running the same or comparable speed / volts?

    I'm mining at 27MH/s on R9 Fury 1000MHz -78mv -30% power and its dead silent 33% fan 69*C. Need to buy power meter this week to check :/
  • wirelessnet2wirelessnet2 Member Posts: 385 ✭✭✭
    Marvell9 said:

    @wirelessnet2 when u buy 50k worth of hardware ..

    o shit ok nvm
  • wirelessnet2wirelessnet2 Member Posts: 385 ✭✭✭
    oh no i just realized somethng... Unless Nvidia Pascal mines ETH better with newer WDDM (1+) drivers, then the GTX 1080 will have shit ETH performance
  • wirelessnet2wirelessnet2 Member Posts: 385 ✭✭✭
    chrizzo said:

    anyone some idea? My 390X get only 28M/h, but the most have 30-35 M/h? Using Win7 x64, crimson 15.12.

    really? lol my R9 390s are getting 31 MH/s. Is it thermal throttling? Whats the ASIC Quality?
  • chrizzochrizzo GermanyMember Posts: 131 ✭✭
    edited May 2016

    chrizzo said:

    anyone some idea? My 390X get only 28M/h, but the most have 30-35 M/h? Using Win7 x64, crimson 15.12.

    really? lol my R9 390s are getting 31 MH/s. Is it thermal throttling? Whats the ASIC Quality?
    No thermal are good.

    On first rig is: 71-78%
    On second rig is: 73-76%
  • workwork Member Posts: 2,084 ✭✭✭✭
    @chrizzo that hashrate definitely seems low. What miner software are you using? What is the full cmdline you run? Are you overclocking?
  • Mo35Mo35 Member Posts: 43
    edited May 2016
    good news people (for thos who still kicking around with cuda 6.5) , had a bit of free time on hands and compiled latest 1.0.8 for us

    http://wikisend.com/download/301718/ethminer-0.9.41-genoil-1.0.8-cuda6.5.zip

    compiled with all features enabled , links TTL is 90 days
  • chrizzochrizzo GermanyMember Posts: 131 ✭✭
    work said:

    @chrizzo that hashrate definitely seems low. What miner software are you using? What is the full cmdline you run? Are you overclocking?

    My miner is ethminer genoil 1.0.8.

    CMD Line:

    setx GPU_SINGLE_ALLOC_PERCENT 100
    set GPU_SINGLE_ALLOC_PERCENT = 100
    ethminer --farm-recheck 200 -G -F http://127.0.0.1:8080/rig1 --cl-extragpu-mem 0

    The cards are now a little bit oced, but without oc the same hashrate.
  • wirelessnet2wirelessnet2 Member Posts: 385 ✭✭✭
    i am a full fan of the claymore miner. No more DAG files... such a smart guy
  • wirelessnet2wirelessnet2 Member Posts: 385 ✭✭✭
    and the dual mining mode is awesome. The Decred I dual mine pays for all the electricity I have to use, and gives a little bit of profit
  • ethfanethfan Member Posts: 458 ✭✭✭
    edited May 2016
    @Genoil, keep getting this at random intervals causing mining to stop...
    ✘ 16:00:00|stratum Read response failed: Operation canceled
    ℹ 16:00:00|stratum Reconnecting in 3 seconds...
    ℹ 16:00:03|stratum Connecting to stratum server eu1.ethermine.org:4444
    terminate called after throwing an instance of 'boost::exception_detail::clone_impl< boost::exception_detail::error_info_injector< boost::system::system_error> >'
    what(): close: Bad file descriptor
    Aborted
    Previously rigs run for days/weeks without stopping. What can be causing this?

    Anyone else with similar issues?
  • GenoilGenoil 0xeb9310b185455f863f526dab3d245809f6854b4dMember Posts: 769 ✭✭✭
    https://github.com/Genoil/cpp-ethereum/issues/38. Is the work timeout also showing in your logs? If yes, you can disable the timeout for now by setting a really high value for it.(i.e --work-timeout 1000) If you don't know, you can try intentionally setting it too low (i.e --work-timeout 10) to see if the work-timeout works or actually triggers this bug.

    It seems a Linux issue with boost::asio/sockets, because I haven't seen it on Windows. I also can't replicate it in order to fix it.
  • LogicaluserLogicaluser Member Posts: 214 ✭✭
    edited May 2016
    @ethfan @Genoil

    I have what may be a related issue on one of my rigs.


    Switched my rigs over to 1.0.8 yesterday afternoon following the DAG transition, and one rig has stopped mining with this error twice since (all the other rigs are fine so far)
    This particular miner was previously stable for weeks other than typical DAG transition crashes.

    Each time it looks like it repeated that error for several hours.... and then oddly enough somehow "fixed" itself when I teamviewer'ed in.... began mining again without me doing anything.

    Rig stats:
    three 7950, one 390
    AMD dual core, 4GB RAM
    Win7 x64 & AMD 15.7.1 drivers
    1050w PSU

    BAT:

    ethminer -S us2.ethpool.org:3333 -FS us1.ethpool.org:3333 -O 0xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx.xxxxx -G --farm-recheck 2000 --farm-retries 0 --opencl-platform 0 --cl-global-work 16384 --cl-local-work 128
    22.png 71.3K
  • GenoilGenoil 0xeb9310b185455f863f526dab3d245809f6854b4dMember Posts: 769 ✭✭✭
    @Logicaluser "could not resolve host" doesn't look like a miner problem, rather a DNS issue of your connection. Did that PC have any other internet traffic while this happened?
  • ethfanethfan Member Posts: 458 ✭✭✭
    @Genoil, so I set --work-timeout 10 and soon saw timeouts happening. However the failover kicked in and the miner just kept going. So, when abort happens there must have been something else. Now going to go the other extreme with --work-timeout 10000. See how long before an abort happens...

    For the failover I am using -FS which is working as I said. I suppose you can have only one -FS? When is -FO used?

    Btw, just love the A0+0:R1+0:F0. Shows exactly what is happening. I have modified the log output to have a nice compact display that only needs 70 columns to show all.

    Hope you can sort out this bug. Hate to roll back to 107.
  • LogicaluserLogicaluser Member Posts: 214 ✭✭
    edited May 2016
    @Genoil the thing is... it's a headless rig accessed by teamviewer
    hence my confusion when I got the offline notification from pool, yet was able to connect and saw hours of that network error... only for it to resolve itself and go right back to mining within moments of connecting, on both occasions

    doesn't appear to have dropped network connection based on router or wireless adapter logs, nor anything odd in event viewer. It's also on the same network as other similar rigs that have ran with 1.0.8 past 24hr without issue.
    It seems this rig previously ran 1.0.3 and qtminer, will do further experimentation to see if I can narrow down a cause
  • excavateexcavate Member Posts: 7
    I had version 1.08 compiled on Ubuntu 14.04 without a problem. However, something became corrupted so I decided to install Ubuntu 16.04 and put genoil miner on there. Was unsuccessful. Read that 16.04 did away with fglrx for AMD cards so I figured that was the issue. Formatted and installed 15.10 and now having this issue below. I am totally at a loss here. Anyone have any advice?

    This is result from lspci and I have the catalyst driver installed. It is detecting R9 290, however, my cards are R9 390?

    00:02.0 VGA compatible controller: Intel Corporation Xeon E3-1200 v3/4th Gen Core Processor Integrated Graphics Controller (rev 06) 01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Hawaii PRO [Radeon R9 290] (rev 80) 05:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Hawaii PRO [Radeon R9 290] (rev 80)



    [ 61%] Generating OpenCL Kernel Byte Array Scanning dependencies of target ethash-cl [ 64%] Building CXX object libethash-cl/CMakeFiles/ethash-cl.dir/ethash_cl_miner.cpp.o [ 66%] In file included from /home/me/cpp-ethereum/libethash-cl/ethash_cl_miner.cpp:40:0: /home/me/cpp-ethereum/libethash-cl/ethash_cl_miner.h:12:21: fatal error: CL/cl.hpp: No such file or directory compilation terminated. libethash-cl/CMakeFiles/ethash-cl.dir/build.make:59: recipe for target 'libethash-cl/CMakeFiles/ethash-cl.dir/ethash_cl_miner.cpp.o' failed make[2]: *** [libethash-cl/CMakeFiles/ethash-cl.dir/ethash_cl_miner.cpp.o] Error 1 CMakeFiles/Makefile2:242: recipe for target 'libethash-cl/CMakeFiles/ethash-cl.dir/all' failed make[1]: *** [libethash-cl/CMakeFiles/ethash-cl.dir/all] Error 2 make[1]: *** Waiting for unfinished jobs....

    I'm open to any suggestions... this is really getting frustrating.
  • dlehenkydlehenky Member Posts: 2,249 ✭✭✭✭
    @excavate Perhaps go back to 14.04, where everything was working?
  • excavateexcavate Member Posts: 7
    @diehenky Thanks, that was going to be my last resort if I couldn't get this worked out.
  • GenoilGenoil 0xeb9310b185455f863f526dab3d245809f6854b4dMember Posts: 769 ✭✭✭
    CL/cl.hpp missing..will bring it back ASAP.
Sign In or Register to comment.