Topic: Claymore's ZCash/BTG AMD GPU Miner v12.6 (Windows/Linux) - page 192. (Read 3839201 times)

with a properly strapped bios the 8pin power was my bottleneck lol.. like 12g wire up in smoke.

you can squeeze 360h/s out of a 290.

What is the power usage that others are getting on 2x AMD 280x GPU's mining with the latest miner ? I know newers card will have better energy use but what is the average for this question ?

275h/s - windforce 290 @1000core/1250mem/w7 and 10/15.12 and 16.11*
295h/s - reference 290 @1000core/1250mem/w7/15.12
335h/s - reference 290w/390 bios @1000core/1250mem/w7/15.12

*elpida mem likely bottlenecking

Hello

Someone use V 11.1 on R9 290? Whats hashrate get?

Thanks.

yeah I think you can.  You can go I believe in advanced system settings, system properties / advanced tab.  then its the environment variables button.  You can see everything from there.

I'd prefer temporary environmental variables too. You don't happen to know what the default is, do you? XD (I already used the setx commands)

Edit: Nevermind, I found that they are stored in the registry and did a quick comparison with another pc. I think you can just delete the entries without any issues

that means is not worth $800. ty

no, they will not reset. the values persist through reboots with SETX. with SET it will only be used in that CMD (or .bat) session and be reset when it closes or the machine reboots.

i prefer environmental variables to only be used for the duration of that mining session in case some other program i will fire up at some point uses those values and behaves unexpectedly. not that i know of any but you never know.

Ok thanks! Will the settings, that were defined with setx, be reset to default upon restart of the pc?

Just for your information, you can put set instead of setx, basically the difference is set is changed only for the duration of the program.  setx is basically a universal change.  Good to see your gpu's are working, sometimes its the little things...

Thanks a lot! I was missing the "setx" I tried with -di 01 (it was not a typo) and it works with both cards now!! It's strange though, I think it worked with the space in between before, at least the GPU's were recognized.
Anyway, I happy I got a working solution, not going to question the logic behind it. Thank you so much!
[/quote]
Great have a great day

Thanks a lot! I was missing the "setx" I tried with -di 01 (it was not a typo) and it works with both cards now!! It's strange though, I think it worked with the space in between before, at least the GPU's were recognized.
Anyway, I'm happy I got a working solution, not going to question the logic behind it. Thank you so much!

First I hope this was just a typo but -di  should be -di 01 not -di 0 1 so far as I know
The set commands go in the start.bat file not the config.txt

setx GPU_FORCE_64BIT_PTR 1
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
ZecMiner64.exe

Alright I reinstalled the 15.12 driver using the lastest version of DUU for clean up. I did several tests for each combination. GPU-Z to see if OpenCL support was recognized. LuxMark as an OpenCL benchmark/test. Compubench for a second OpenCL test and finally I tested the v8 and v11 miners. Just a quick remark: the Compubench test is really strange and crashed everytime I had only 1 card installed. No idea what to make of that, maybe just a poorly written benchmark tool.
Here are the results:

Both cards - GPU 1 in 1st slot & GPU 2 in 2nd slot:
GPU-Z, LuxMark, Compubench were all positive.
v11 miner worked only for the GPU1 (in 1st slot) card e.g for -di 1. For -di  1 0 it would hang and the console window wouldn't even close. Shutting down the pc lead to a boot loop, only hard reset helped.
v8 miner also worked only for the GPU1.

GPU1 in 1st slot:
GPU-Z & LuxMark passed, Compubench crashed
v11: passed
v8: passed

GPU1 in 2nd slot:
GPU-Z & LuxMark passed, Compubench crashed
v11: passed
v8: passed

GPU2 in 1st slot:
GPU-Z & LuxMark passed, Compubench crashed
v11: crashed with with Hashrate not equal to 0, crashed after 2 submitted shares
v8: crashed with H/s = 0 and caused the to driver reset

GPU2 in 2nd slot:
GPU-Z & LuxMark passed, Compubench crashed
v11: crashed with H/s = 0 and "OC v5, Reset control for GPU0" in the log file
v8: crashed with H/s = 0 and "OC v5, Reset control for GPU0" in the log file

Both GPUs in reversed order:
GPU-Z, LuxMark, Compubench were all positive.
v11 miner worked only for the GPU1 (in 2nd slot) card e.g for -di 0. Using -di 1 caused the driver to crash
v8 miner worked only for the GPU1 (in 2nd slot) card e.g for -di 0. Using -di 1 caused the driver to crash

Sorry for the long post, but this behavoir is really unexplainable to me, especially since it worked just fine before the driver update
Also another question: How do you parse the GPU_MAX_ALLOC_PERCENT 100 to the program? If I write it in the config file it returns a red error message, telling me it's an unknown command. And everything I write into the .bat file is ignored somehow.

Thanks a lot for your willingness to help me so far!

 Due to lower clocks they should be a little slower, but close, to the Consumer version with the same GPU chip.

OK. You still don't know how miner clamor can run in dual mode eth+
can't paste the picture in
Claymore best of all

thnx a lot

I use the latest version of the miner without settings in config file. Driver for AMD latest, control cooling through MSI Afterburner. Memory 1700 GPU 1080  voltage of -50 mV using the AMD driver Wattman. 200 s\h stable.

this post has actually started me tinkering with a freebsd box again now that amdgpu is a bit more seasoned. i'm confident i can get something going and will report back.. eventually.

alternatively DragonflyBSD and OpenBSD look like interesting prospects with surprising compatibility.

I would try the two cards one at a time in slot 2.  Also do you have any other card that you can use in slot 2 to see if it works. And how is  -di set.