Author

Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 584. (Read 3426922 times)

sr. member
Activity: 350
Merit: 250
sooo, who's up for doing some profit calculations for us?  Grin
groestl is now no longer profitable after the huge amount of people that jumped onto it when the price shot up.
right now with 3 750Ti's, and a power usage of only 155W for the entire system, daily profit is £0.08 or $0.12

sooo, profit calcs anyone?
member
Activity: 80
Merit: 10
Just wondering, has anyone tried the latest beta linux drivers that enable overclocking? I'm using my 750Tis in my media/backup server, so it's running linux. Would be good to boost the hashrate a little!

Also, any documentation yet for the latest commits on Git? I was wondering how the failover works i.e. does it return to the original pool when it comes back up? I'm not using it yet as an mining YAC and I believe the latest don't do jane yet?
sr. member
Activity: 350
Merit: 250
Is there key to stop cudaminer/ccminer?
Asking because killing window of the program doesn't seem to work flawlessly every time
Last night I start/stopped several times and PC started working erratically - altho no miner was working the card (750Ti) was working in 3D mode and I had to restart.


CTRL+C
legendary
Activity: 1151
Merit: 1001
Is there key to stop cudaminer/ccminer?
Asking because killing window of the program doesn't seem to work flawlessly every time
Last night I start/stopped several times and PC started working erratically - altho no miner was working the card (750Ti) was working in 3D mode and I had to restart.
legendary
Activity: 1400
Merit: 1050
What version of cudaminer is it i keep seeing? Being every screen shot lately reports the users temps?

It's the latest compile from source files on github. There was an update on github ~2 days ago. If you compiled from that you should see the new temp readings. That said, the latest compile is a little slower for me compared to the previous version (scrypt-N at least, haven't tested other algos).

The NVML api only works with the 64 bit cudaminer version on windows. nVidia doesn't provide a 32 bit NVML DLL.

Now unfortunately the 64 bit cudaminer mines a bit slower in scrypt and scrypt-jane... I wonder if it is possible to combine 32 bit CUDA device code with a 64 bit host code?

most likely, most of us will be running 64 bit windows anyway so the 32bit code should run the same as normal, and the 64bit run as its meant to. you would have to try it as i am unable to compile 64 bit due to visual c++ express Sad
usually I have better result with x86 (to the level I usually don't compile the x64 version)
I believe too some big hack is possible at this level to make the api work with the x86 version of cudaminer... however it won't work with native windows x86 (in case some are still using it...)

I saw an email saying that Cuda 6 was out, anything to gain using it, or will it be like the RC ?
sr. member
Activity: 350
Merit: 250
The NVML api only works with the 64 bit cudaminer version on windows. nVidia doesn't provide a 32 bit NVML DLL.

Now unfortunately the 64 bit cudaminer mines a bit slower in scrypt and scrypt-jane... I wonder if it is possible to combine 32 bit CUDA device code with a 64 bit host code?

most likely, most of us will be running 64 bit windows anyway so the 32bit code should run the same as normal, and the 64bit run as its meant to. you would have to try it as i am unable to compile 64 bit due to visual c++ express Sad

i think i may have to look into NVML api. right now i am trying to use nvapi and it seems buggy as sh*t
hero member
Activity: 756
Merit: 502
What version of cudaminer is it i keep seeing? Being every screen shot lately reports the users temps?

It's the latest compile from source files on github. There was an update on github ~2 days ago. If you compiled from that you should see the new temp readings. That said, the latest compile is a little slower for me compared to the previous version (scrypt-N at least, haven't tested other algos).

The NVML api only works with the 64 bit cudaminer version on windows. nVidia doesn't provide a 32 bit NVML DLL.

Now unfortunately the 64 bit cudaminer mines a bit slower in scrypt and scrypt-jane... I wonder if it is possible to combine 32 bit CUDA device code with a 64 bit host code?

sr. member
Activity: 350
Merit: 250
I am unsure if it is a linux temp checker or if it uses nvapi. I think nvapi works in linux aswell but the usual command line chrcker for linux wont run on windows

Ahh ok thanks DrAlco. I didnt rralise christian had put it in, i was about to put my own in  Roll Eyes
newbie
Activity: 43
Merit: 0
What version of cudaminer is it i keep seeing? Being every screen shot lately reports the users temps?

It's the latest compile from source files on github. There was an update on github ~2 days ago. If you compiled from that you should see the new temp readings. That said, the latest compile is a little slower for me compared to the previous version (scrypt-N at least, haven't tested other algos).
legendary
Activity: 1400
Merit: 1050
It is either one someone else made or a new release
It doesn't seem to be in the binary I put here last week for testing: it says "NVML GPU monitoring is not available" (or something is missing...)
sr. member
Activity: 350
Merit: 250
It is either one someone else made or a new release
full member
Activity: 146
Merit: 100
What version of cudaminer is it i keep seeing? Being every screen shot lately reports the users temps?

I believe it is a flag actually, that gives this kind of message. I tried the debug command, but that did not give that information, so I'm not so sure anymore. Perhaps it is indeed a newer version, and that does have the appropriate command flag.
sr. member
Activity: 350
Merit: 250
What version of cudaminer is it i keep seeing? Being every screen shot lately reports the users temps?
full member
Activity: 161
Merit: 100
Christian!,

I'm getting about an 3kh/s hashrate increase per card with your latest github source compared to the last release!  Seems to be moving in the right direction;  I see they just released Cuda Toolkit 6.0, any thoughts if it offers any improvements or would even work?

(Scrypt-N)

legendary
Activity: 1400
Merit: 1050
Windows 8.1 sucks, and will always suck: have you tried a black/transparent theme? Yeah the text at the top of your window is near invisible. It is fast though.
As for the cards sucking a load of extra power, is it caused by some risers maybe?
Hell I am sad about that €302 I sold my bitcoin for now, but I am NOT buying back like a fool Roll Eyes
is it going up again ? yeah !!! (I didn't sell yet the Max Keiser Fund... ) I stopped checking the rate when it started to become too depressing
full member
Activity: 182
Merit: 100
Windows 8.1 sucks, and will always suck: have you tried a black/transparent theme? Yeah the text at the top of your window is near invisible. It is fast though.
As for the cards sucking a load of extra power, is it caused by some risers maybe?
Hell I am sad about that €302 I sold my bitcoin for now, but I am NOT buying back like a fool Roll Eyes
sr. member
Activity: 350
Merit: 250
Sli im not sure. Their dual fans, and no external power. The msi cards seems to be the best overclockers

Ive found the code to get the gpu temp from the nvidia api, so tomorrow i will be going through it, finding how to get device name, temperature, and fan speed from the device id sent by -d and then output those details through the webserver aswell. Just a small update for anyone interested
hero member
Activity: 676
Merit: 500
Ahh ok, my msi cards handle up to 1400MHz on stock bios
This with ccminer5 I guess , My cards are with one fan only and they are small formfactor and they don`t have external power connector.  Smiley Man does your card have sli , i thought ti750 don`t have this option?
The msi site has a misleading picture of twinfrost ,  on a sli card ,
sr. member
Activity: 350
Merit: 250
Ahh ok, my msi cards handle up to 1400MHz on stock bios
legendary
Activity: 1292
Merit: 1000
Today version from github...

There are some unrealistic values of hashrate per GPU at the begining, for example:
GPU #1: GeForce GTX 750 Ti, 7837 khash/s (Vertcoin)
GPU #0: GeForce GTX 750 Ti, 5478 khash/s
Smiley

After few minutes is already OK.



Code:
Apr 15 23:19:43 kopiemtu cpuminer[6100]: Starting Stratum on stratum+tcp://coinotron.com:3340
Apr 15 23:19:43 kopiemtu cpuminer[6100]: NVML GPU temperature, fan, power monitoring enabled.
Apr 15 23:19:43 kopiemtu cpuminer[6100]: 2 miner threads started, using 'scrypt' algorithm.
Apr 15 23:19:43 kopiemtu cpuminer[6100]: Stratum detected new block
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #1: GeForce GTX 750 Ti with compute capability 5.0
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #1: interactive: 0, tex-cache: 0 , single-alloc: 0
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #1: 32 hashes / 8.0 MB per warp.
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #0: GeForce GTX 750 Ti with compute capability 5.0
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #0: interactive: 0, tex-cache: 0 , single-alloc: 0
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #0: 32 hashes / 8.0 MB per warp.
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #1: Performing auto-tuning (Patience...)
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #0: Performing auto-tuning (Patience...)
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #1: maximum total warps (BxW): 245
Apr 15 23:20:11 kopiemtu cpuminer[6100]: GPU #0: maximum total warps (BxW): 245
Apr 15 23:21:13 kopiemtu cpuminer[6100]: GPU #0: 153501.76 hash/s with configuration T5x24
Apr 15 23:21:13 kopiemtu cpuminer[6100]: GPU #0: using launch configuration T5x24
Apr 15 23:21:13 kopiemtu cpuminer[6100]: GPU #0: GeForce GTX 750 Ti, 5478 khash/s
Apr 15 23:21:13 kopiemtu cpuminer[6100]:         Temp:  54C  Fan speed:  35%  Power:  N/A
Apr 15 23:21:14 kopiemtu cpuminer[6100]: GPU #1: 155947.53 hash/s with configuration T5x24
Apr 15 23:21:14 kopiemtu cpuminer[6100]: GPU #1: using launch configuration T5x24
Apr 15 23:21:14 kopiemtu cpuminer[6100]: GPU #1: GeForce GTX 750 Ti, 7837 khash/s
Apr 15 23:21:14 kopiemtu cpuminer[6100]:         Temp:  57C  Fan speed:  37%  Power:  N/A
Apr 15 23:21:45 kopiemtu cpuminer[6100]: GPU #1: GeForce GTX 750 Ti, 146.73 khash/s
Apr 15 23:21:45 kopiemtu cpuminer[6100]:         Temp:  59C  Fan speed:  38%  Power:  N/A
Apr 15 23:21:45 kopiemtu cpuminer[6100]: accepted: 1/1 (100.00%), 5625 khash/s (yay!!!)
Apr 15 23:21:53 kopiemtu cpuminer[6100]: GPU #1: GeForce GTX 750 Ti, 148.05 khash/s
Apr 15 23:21:53 kopiemtu cpuminer[6100]:         Temp:  60C  Fan speed:  38%  Power:  N/A
Apr 15 23:21:53 kopiemtu cpuminer[6100]: accepted: 2/2 (100.00%), 5626 khash/s (yay!!!)
Apr 15 23:22:01 kopiemtu cpuminer[6100]: GPU #1: GeForce GTX 750 Ti, 147.99 khash/s
Apr 15 23:22:01 kopiemtu cpuminer[6100]:         Temp:  60C  Fan speed:  38%  Power:  N/A
Apr 15 23:22:01 kopiemtu cpuminer[6100]: accepted: 3/3 (100.00%), 5626 khash/s (yay!!!)
Apr 15 23:22:32 kopiemtu cpuminer[6100]: GPU #1: GeForce GTX 750 Ti, 146.73 khash/s
Apr 15 23:22:32 kopiemtu cpuminer[6100]:         Temp:  61C  Fan speed:  38%  Power:  N/A
Apr 15 23:22:33 kopiemtu cpuminer[6100]: accepted: 4/4 (100.00%), 5625 khash/s (yay!!!)
Apr 15 23:23:07 kopiemtu cpuminer[6100]: Stratum detected new block
Apr 15 23:23:07 kopiemtu cpuminer[6100]: GPU #1: GeForce GTX 750 Ti, 146.85 khash/s
Apr 15 23:23:07 kopiemtu cpuminer[6100]:         Temp:  62C  Fan speed:  39%  Power:  N/A
Apr 15 23:23:07 kopiemtu cpuminer[6100]: GPU #0: GeForce GTX 750 Ti, 145.13 khash/s
Apr 15 23:23:07 kopiemtu cpuminer[6100]:         Temp:  58C  Fan speed:  37%  Power:  N/A
Apr 15 23:24:07 kopiemtu cpuminer[6100]: GPU #0: GeForce GTX 750 Ti, 144.92 khash/s
Apr 15 23:24:07 kopiemtu cpuminer[6100]:         Temp:  58C  Fan speed:  37%  Power:  N/A
Apr 15 23:24:07 kopiemtu cpuminer[6100]: GPU #1: GeForce GTX 750 Ti, 146.38 khash/s
Apr 15 23:24:07 kopiemtu cpuminer[6100]:         Temp:  63C  Fan speed:  39%  Power:  N/A

Jump to: