Pages:
Author

Topic: DiabloMiner GPU Miner - page 82. (Read 866206 times)

Rai
newbie
Activity: 12
Merit: 0
December 29, 2010, 09:28:22 AM
-w does not control lag in that way. It also needs to be a power of 2, and depending on your card, the maximum -w is either 256 or 512. Tuning it to optimum speed is all you can do with it, it won't stop lag.

Good to know that.

I suspect nvidia drivers are terminally braindead. Make sure you're using the newest version. Also, 72 mhash > 66. Cheesy

Yes, but being able to use my computer while mining trumps not being able to, so 6 mhash is the price I pay for that.
legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 29, 2010, 04:55:03 AM
-w does not control lag in that way. It also needs to be a power of 2, and depending on your card, the maximum -w is either 256 or 512. Tuning it to optimum speed is all you can do with it, it won't stop lag.

I suspect nvidia drivers are terminally braindead. Make sure you're using the newest version. Also, 72 mhash > 66. Cheesy
Rai
newbie
Activity: 12
Merit: 0
December 28, 2010, 06:13:22 PM
Try using -f 1000, this seems to fix the overall desktop interactivity problem, but loses performance.

Unfortunately, not in this case.  The performance did not drop, but the bogdown is still just as terrible.  I'll try it on my other computer though, that one is not Hybrid.  I'll let you know how it turns out.

EDIT:

My other computer has two GTS 250's, with an SLI setup.  With m0mchil's miner, I have to run two separate processes, and each clocks in around 33000 khps.  With DiabloMiner, I have to only run one process and I get approximately 72000 khps, but again there's a terrible bogdown.  Playing with -w and using -f 1000 again does not drop performance, but doesn't help either.  Sad  Oh well, looks like I have to stick with m0mchil's miner on both machines.
legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 28, 2010, 05:18:27 PM
BTW, the java is using 100% CPU in my computer, not 0% as advertised. I want my money back! Wink

I'm using the last version. Do you want me to test anything?

It should only use 100% CPU for like 15-30 seconds, and then cut down to almost none.

I did a comparison with m0mchil's miner and this one.  I got both to work, but they produce the same khps rate, averaging right around 21000 khps.  This is on a laptop, with an nVidia GTX 260M Hybrid SLI setup (with a GeForce 9400, but it only recognizes the GTX, latest drivers installed), running Windows 7 Ultimate.  Only difference is that there's a horrible slowdown when using the DiabloMiner, and no processes are running at 100%.  I'm not sure what it is that's bogging down the system, but it only happens when running DiabloMiner, not m0mchil's.  On a positive note, I was able to run minerd alongside DiabloMiner without any khps loss on either process.  I've toyed with with the -g -w and -f switches, but nothing really seems to help.  I suppose this particular miner isn't all that great for folks running nVidia hardware.  Unless there are some better suggestions to improve the performance of this thing... other than the usual "get an ATI/get linux" responses.

Hybrid is generally broken on NVidia hardware; OSX has it worse than Windows. It seems in Windows it either only recognizes the first GPU or the current one (and won't switch while running either way). I don't suggest mining on hybrid machines until the manufacturers figure it out.

Also, NVidia users tend to have issues with OpenCL apps in general due to poor design in the driver. Try using -f 1000, this seems to fix the overall desktop interactivity problem, but loses performance.
legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 28, 2010, 05:14:42 PM
Update: Reverted global getwork instance work, it was causing more problems than
it was fixing. Increased khash meter accuracy. Removed stuck executor
thread code, this didn't work right and the thread eventually unsticks
itself on applicable platforms anyhow.
Rai
newbie
Activity: 12
Merit: 0
December 28, 2010, 12:14:41 PM
I did a comparison with m0mchil's miner and this one.  I got both to work, but they produce the same khps rate, averaging right around 21000 khps.  This is on a laptop, with an nVidia GTX 260M Hybrid SLI setup (with a GeForce 9400, but it only recognizes the GTX, latest drivers installed), running Windows 7 Ultimate.  Only difference is that there's a horrible slowdown when using the DiabloMiner, and no processes are running at 100%.  I'm not sure what it is that's bogging down the system, but it only happens when running DiabloMiner, not m0mchil's.  On a positive note, I was able to run minerd alongside DiabloMiner without any khps loss on either process.  I've toyed with with the -g -w and -f switches, but nothing really seems to help.  I suppose this particular miner isn't all that great for folks running nVidia hardware.  Unless there are some better suggestions to improve the performance of this thing... other than the usual "get an ATI/get linux" responses.
newbie
Activity: 24
Merit: 0
December 28, 2010, 02:39:37 AM
BTW, the java is using 100% CPU in my computer, not 0% as advertised. I want my money back! Wink

I'm using the last version. Do you want me to test anything?
legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 25, 2010, 01:39:15 PM
For those wondering, I replied to the bug. OSX hybrid graphics are screwing up OpenCL apps.
newbie
Activity: 12
Merit: 0
December 25, 2010, 12:22:54 PM
I have tried DiabloMiner on MacOS X but I have some errors when DiabloMiner is creating the command queue:

[12/25/10 5:18:52 PM] Started
[12/25/10 5:18:52 PM] Added GeForce 9400M (#1) (2 CU, local work size of 256)
[12/25/10 5:18:54 PM] Added GeForce 9600M GT (#2) (4 CU, local work size of 256)
[12/25/10 5:18:54 PM] ERROR: [CL_DEVICE_NOT_AVAILABLE] : OpenCL Error : clCreateCommandQueue failed: Device 0x1020331f0 is not available.
[12/25/10 5:18:54 PM] ERROR: Failed to allocate queue
[12/25/10 5:18:54 PM] ERROR: [CL_INVALID_COMMAND_QUEUE] : OpenCL Error : clEnqueueWriteBuffer failed: invalid command queue 0x0.
Exception in thread "DiabloMiner Executor (GeForce 9600M GT (#2)/0)" java.lang.NullPointerException
        at org.lwjgl.opencl.APIUtil.releaseObjects(APIUtil.java:506)
        at org.lwjgl.opencl.APIUtil.releaseObjects(APIUtil.java:502)
        at org.lwjgl.opencl.CL10.clReleaseCommandQueue(CL10.java:565)
        at com.diablominer.DiabloMiner.DiabloMiner$DeviceState$ExecutionState.run(DiabloMiner.java:568)
        at java.lang.Thread.run(Thread.java:680)

I opened a ticket at github with more information  : https://github.com/Diablo-D3/DiabloMiner/issues#issue/3

Thank you for your work,

legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 24, 2010, 05:29:18 PM
Update: Forced all accesses to a getwork instance to be synchronous, should fix final network thread problems
legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 24, 2010, 05:08:11 PM
Throwing up a dummy xvfb will indeed fix this. ATI users don't have this problem because X has to be running to use OpenCL
newbie
Activity: 59
Merit: 0
December 24, 2010, 07:18:25 AM
The ELFCLASS32 bug is, unfortunately, inside the AWT code inside Java. lwjgl rams into it because it initializes AWT even if I do not open a window. They are aware of the bug; Oracle has no plans on fixing it on their end (ANY app that tries to init AWT on a system without X running will do this).

Smiley

Well, anyway I'd just thought posting up might help others in a situation like mine. I really wanted to get the GPU miner working, but don't want X...

Thanks for all your work on this project!
legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 24, 2010, 07:16:26 AM
The ELFCLASS32 bug is, unfortunately, inside the AWT code inside Java. lwjgl rams into it because it initializes AWT even if I do not open a window. They are aware of the bug; Oracle has no plans on fixing it on their end (ANY app that tries to init AWT on a system without X running will do this).
newbie
Activity: 59
Merit: 0
December 24, 2010, 06:33:51 AM
I found I kept getting errors when running the miner...

I fixed the ELF_CLASS error by deleting the non *64.so libs and renaming them to *.so (just remove the 64!)

Then, I got an error about not being able to connect to a screen (it's a headless server);

I installed xvfb...ran this;

export DISPLAY=:1
Xvfb :1 -screen 0 1024x768x16 &

and it worked fine  Cool

You didn't have to rename the libs. Setting DISPLAY alone fixes this (as long as you have an X server running).

I don't have an X server running, or didn't before xvfb.

Also, it did repeatedly crash before getting as far as crashing about X related things, moaning about the ELF CLASS. Renaming the libs fixed this.
hero member
Activity: 489
Merit: 504
December 22, 2010, 12:53:54 PM
Does anybody have a comparison on how Linux against Windows machines are performing when running the GPU miners?
No, but Linux runs faster.
Great news Cheesy
I was already dreading I'd have to setup a Windows machine ^^

legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 22, 2010, 11:59:59 AM
Does anybody have a comparison on how Linux against Windows machines are performing when running the GPU miners?

No, but Linux runs faster.
hero member
Activity: 489
Merit: 504
December 22, 2010, 08:51:43 AM
Does anybody have a comparison on how Linux against Windows machines are performing when running the GPU miners?
legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 21, 2010, 11:56:35 PM
Update: Made the getwork clone from network thread method more paranoid, should help with low difficulty targets like pools and the testnet.
legendary
Activity: 1162
Merit: 1000
DiabloMiner author
December 21, 2010, 11:55:38 PM
I found I kept getting errors when running the miner...

I fixed the ELF_CLASS error by deleting the non *64.so libs and renaming them to *.so (just remove the 64!)

Then, I got an error about not being able to connect to a screen (it's a headless server);

I installed xvfb...ran this;

export DISPLAY=:1
Xvfb :1 -screen 0 1024x768x16 &

and it worked fine  Cool

You didn't have to rename the libs. Setting DISPLAY alone fixes this (as long as you have an X server running).
newbie
Activity: 59
Merit: 0
December 21, 2010, 08:40:40 PM
I found I kept getting errors when running the miner...

I fixed the ELF_CLASS error by deleting the non *64.so libs and renaming them to *.so (just remove the 64!)

Then, I got an error about not being able to connect to a screen (it's a headless server);

I installed xvfb...ran this;

export DISPLAY=:1
Xvfb :1 -screen 0 1024x768x16 &

and it worked fine  Cool
Pages:
Jump to: