Does cgminer overrule the fan/voltage/etc settings i have done with the MSI-afterburner?
Only if you give it parameters to change any of the settings, otherwise all it does is monitor the values. |
|
|
|
Running RDP to the mining rig kills the GPU and crash's CGMiner and requires a reboot of the whole mess. That's been my experience anyway.
Same experience here. RDP is awful for mining. I've also found tightvnc works best for win 7. Does monitoring work via tightvnc? |
|
|
|
|
Server is being DDoSed so download links are broken again till it's over. Looks like DDoS is the new black.
|
|
|
|
|
Changing SDKs there is no need for a recompile, but you must delete any .bin files generated so they're regenerated. Note there is no clean way to uninstall though so be careful when removing the old one to make sure it really has gone. AMD seems to have taken its ideas for uninstall from windows where you just say it is uninstalled but leave half the files there.
|
|
|
|
As far as I'm aware I should be able to go up to ~20000 TC, what's up?
Any help greatly appreciated.
Any ideas? To update I changed TC to --shaders 1792 (from memory, whatever the 7950 shaders are) and it worked - but was only hashing at about 8 kh/s a card. I'm more than happy to try anything suggested. Cheers. README FAQs Ok. I've read the README again. I mean this respectfully because I know how frustrating it would be when seemingly someone is asking questions that are answered previously or in the README but as far as I can tell my specific problem isn't addressed. I've spent about 4-5 hours trying to get it to work, configuring with multiple APP SDK and Catalyst versions, read through the README and SCRYPT-README multiple times, but still can't seem to solve this issue. I'm usuing Xubuntu 12.10, Catalyst 13.1, AMD ADL SDK 5.0, AMD APP SDK 2.8. I will provide the full debug of the mining when I get home (at work). I have 8GB of ram as I know LTC mining is RAM intensive. |
|
|
|
As far as I'm aware I should be able to go up to ~20000 TC, what's up?
Any help greatly appreciated.
Any ideas? To update I changed TC to --shaders 1792 (from memory, whatever the 7950 shaders are) and it worked - but was only hashing at about 8 kh/s a card. I'm more than happy to try anything suggested. Cheers. README FAQs |
|
|
|
Quick question: is it possible to mine with the stratum protocol through an http proxy?
Theoretically yes, but only if the http proxy is very permissive, and none will by default. A socks proxy will definitely work though. |
|
|
|
|
Nope, it sucks dicks, as the cgminer readme says.
|
|
|
|
|
In your initial post it's clear you were mining on one device only, with multiple opencl instances.
SDK 2.8 sucks dicks (on linux at least), as do many of the drivers after the 11.12 prerelease driver with GCN support (which was called amd-driver-installer-8.921-x86.x86_64.run on linux for example), so I guess 12.8 is your next best bet. Uninstalling the SDK is non-trivial but you should get rid of 2.8 and try 2.7. Then make sure all .bin files are deleted so that cgminer recreates new ones with the changed SDK.
Bear in mind that some motherboards may appear to support multiple GPUs, but do NOT actually run, or barely run. My GD70 based rig would only run with 3 GPUs with the default Xorg set up using a login manager and refuse to even start with the 4th GPU in there unless I ran an absolutely stripped down Xorg with just startx us a sole user on xfce and no login manager. Usual set up of drivers is required for all the GPUs with the force option and any distribution provided drivers are a nono.
Yes folks, mining is non-trivial and scrypt mining is yet much more non-trivial. Make sure if you are building a rig that you are using hardware and software that is already known to work as others have tested on these forums before. That can be hard...
|
|
|
|
Hi, I tried to compile cgminer on OSX 10.6.8 with scrypt option to mine litecoin but when i start it I get this error /System/Library/Frameworks/OpenCL.framework/Resources/cl_kernel.h:2195:29: note: instantiated from:
#define rotate(X,Y) __CLFN_I2(X,Y,rotate)
^
/System/Library/Frameworks/OpenCL.framework/Resources/cl_kernel.h:22:26: note: instantiated from:
#define __CLFN_I2(x,y,R) __builtin_overload(2, x, y,\
^~~~~~~~ [2013-04-11 09:45:13] Failed to init GPU thread 0, disabling device 0
[2013-04-11 09:45:13] Restarting the GPU from the menu will not fix this.
[2013-04-11 09:45:13] Try restarting cgminer. any hint to fix it? Nope. Doesn't look like the scrypt opencl kernel compiles on the OSX opencl implementation. You're screwed. |
|
|
|
solo mining = hashes accepted only if it matches the network difficulty, which will take forever.
Ah, that makes sense... going to be a long while never before I see an "Accepted" then  There, FTFY. |
|
|
|
Ah the life of a Pool Op. People only see the 3 or 5% and think they're getting passively rich. Being a big time Pool Op is more or less a full time job.
Full time job would be 40 hours/week (US). I'd kill to only work that many hours on my pool, and I'm sure Graet is the same way I was curious about starting a pool just for fun. This thread changed my mind, it doesn't sound fun at all.
I wasn't joking about: The benefits? Realistically: many sleepless nights, headaches, heartaches, trolling, frustration, loss of money and the very slim chance of making a profit...
and eleuthria with: Con: Dealing with botnets that screw up your servers by overloading them and then dealing with assholes who DDoS you when you kick them off.
Eleuthria is making a sizeable profit now but that belies just how much investment and work he has and is continually putting into it. |
|
|
|
|
Working hardware also needs working software. Don't make the mistake other manufacturers did the first time around and ignore this, trying to do it in house or engaging developers almost after releasing the hardware. Cgminer hashes more of the bitcoin network than any other software and I implore you to engage the developers to get full high quality support and performance for your hardware. Engaging us early with design and protocol decisions and providing hardware to kanoi and myself to develop working software of an appropriate standard is also a guaranteed way of getting people to believe you on these forums.
|
|
|
|
|
Can't remote into windows and start with monitoring. Must start cgminer locally on the machine for monitoring to work.
|
|
|
|
CFLAGS="-o2 -Wall -march=native" ./configure
Take a reaaaaally good look at that and the readme. |
|
|
|
|
See now why would you want a self moderated topic when you can have trolls coming in making the thread so much more fun?
|
|
|
|
|
cgminer -n
will tell you the platforms and devices installed, then select one with
--gpu-platform
|
|
|
|
|
Alas the p2pool end is not exactly playing ball with the avalons though, and I'm not touching the p2pool code itself, just trying to make Avalons work well and be prepared to work with p2ool design, assuming p2pool scales and does all the right things. The current p2pool git master does not meet that criteria, and Aseras himself has been playing with that as you can see above.
|
|
|
|
|
Show Posts
