I'm using, with mixed results. It takes a fair bit of tuning, starting with the default intensity being way too low (it should be closer to 25). On my Windows machine a GTX 980 with ccminer went from 1.9 Ghash/s to 2.0 Ghash/s. But on a linux box wit the same GPU, I couldn't match the hash rate of ccminer. ::shrug:: On yet another Linux machine with AMD R9 GPUs, it was just a hair faster than cgminer after much tuning. However, it's less buggy than cgminer 3.7.2. I can actually hit q to quit without a crash on one machine with this trouble. ADL seems to work much better. Oh, and maybe I just missed an option, but I had to run a dcrd instance with --notls to connect sgminer.
I am having problems with the most basic run of commands already. And when using ./sgminer -n this is output: Code: sgminer -n [12:27:31] CL Platform vendor: Advanced Micro Devices, Inc. [12:27:31] CL Platform name: AMD Accelerated Parallel Processing [12:27:31] CL Platform version: OpenCL 2.0 AMD-APP (1800.11) [12:27:31] Platform devices: 1 [12:27:31] 0 Capeverde [12:27:31] ATI ADL Overdrive5 API found. [12:27:31] ATI ADL Overdrive6 API found. [12:27:31] Found 6 logical ADL adapters [12:27:31] GPU 0 assigned: iAdapterIndex:0 iPresent:1 strUDID:256:26685:4098:57876:5963 iBusNumber:1 iDeviceNumber:0 iDrvIndex:0 iFunctionNumber:0 iVendorID:4098 name:AMD Radeon R7 200 Series [12:27:31] GPU 0 AMD Radeon R7 200 Series hardware monitoring enabled [12:27:31] ADL GPU 0 is Adapter index 0 and maps to adapter id 1622379904 [12:27:31] GPU 0 BIOS partno.: 113-21400XTHE-X07, version: 015.017.000.000.000000, date: 03/01/12 00:50 [12:27:31] Failed to ADL_Overdrive5_FanSpeed_Get for default value [12:27:31] Unable to load the NVIDIA Management Library
I installed OpenCL then the compile worked. But now I have no clue how to adjust my GPU commands in the miner command. It doesnt even start with most basi, looks like this now: Code: ./runmine.sh [11:54:21] Started sgminer 5.3.2-dcr [11:54:21] * using Jansson 2.7 [11:54:21] Probing for an alive pool [11:54:21] Startup GPU initialization... Using settings from pool dcr.suprnova.cc. [11:54:21] Startup Pool No = 0 [11:54:21] Initialising kernel darkcoin-mod.cl with nfactor 10, n 1024 [11:54:21] Maximum buffer memory device 0 supports says 364904448 [11:54:21] Your settings come to 536870912 [11:54:21] Failed to init GPU thread 0, disabling device 0 [11:54:21] Restarting the GPU from the menu will not fix this. [11:54:21] Re-check your configuration and try restarting. [11:54:21] thread_prepare failed for thread 0 ./runmine.sh: Zeile 6: 15279 Speicherzugriffsfehler ./sgminer -k x11mod -o stratum+tcp://dcr.suprnova.cc:2252 -u legalmatFireCreds.testingGPU -p and345 -I 13
Hi,, I installed all like told here: http://ubuntuforums.org/showthread.php?t=2243387 The script I start I created myself. and this is inside the script: Code: ./sgminer -k x11mod -o stratum+tcp://dcr.suprnova.cc:2252 -u legalmatFireCreds.testingGPU -p and345 -I 13 I will now try with your tips. In the manual sgminer --help they dont specify how to use the decred algo. Ahh, I think I have found: https://github.com/sgminer-dev/sgminer/blob/master/doc/kernel.md
Even easier. In the folder sgminer/kernel are all available kernel files. When it says decred.cl it will be most probably used in command with "decred" also only to get Code: ./sgminer -k decred -o stratum+tcp://dcr.suprnova.cc:2252 -u legalmatFireCreds.testingGPU -p and345 --gpu-memclock 850 --thread-concurrency 8192 -I 18 I thought and it works. Now I can check if it gives activity on suprnova. Thanks for pointing this one out. I had it in front py my nose, but I thought, it would give an explicite error if the algo is wrong. Thought it would try to start anyhow. But it seems to only was this problem. Thanks a lot... But there are easy n00b proof documentations lacking on mining decred. I never had more problems to find my way. ;-)
=Ok, my next step would be then to get it to run on nvidia. When driver setup works it should be pretty mcuh the same steps, correct?
The hasrate with a Radeon 7770 on linux with the settings above is around 415.18 MH/s BTW The GPU is utilized at a 86%. Maybe there are suggestions on improving? Code to check the usage is: Code: aticonfig --odgc --odgt Default Adapter - AMD Radeon R7 200 Series Core (MHz) Memory (MHz) Current Clocks : 1150 1250 Current Peak : 1150 850 Configurable Peak Range : [300-1650] [150-1750] GPU load : 85% Default Adapter - AMD Radeon R7 200 Series Sensor 0: Temperature - 69.00 C