[ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX]

[BlueDragon747]

Would be awesome to see support for Blake-256

more coins are using this efficient algo 

[1714566210]

1714566210

[1714566210]

1714566210

[1714566210]

1714566210

[ivanlabrie]

Gonna try this cool winpe 64 with the cudaminer 64bit binaries...

http://windowsmatters.com/

thoughts?

[chapmanjw]

@cbuchner1: Is it possible to set the "intensity" for CudaMiner like cgminer and other GPU miners?  When I am doing other things on my machine, it would be nice to have a way to lower its resource utilization.

-i 1 = useable pc
-i 0 = somewhat useable pc

Yea, that's only slightly better for me with -i 1.  Still slows things down quite a bit and can't really do work while its mining. 

[Banin]

@cbuchner1: Is it possible to set the "intensity" for CudaMiner like cgminer and other GPU miners?  When I am doing other things on my machine, it would be nice to have a way to lower its resource utilization.

-i 1 = useable pc
-i 0 = somewhat useable pc

Yea, that's only slightly better for me with -i 1.  Still slows things down quite a bit and can't really do work while its mining. 

its what interactive mean ...

[UNOE]

@cbuchner1: Is it possible to set the "intensity" for CudaMiner like cgminer and other GPU miners?  When I am doing other things on my machine, it would be nice to have a way to lower its resource utilization.

-i 1 = useable pc
-i 0 = somewhat useable pc

Yea, that's only slightly better for me with -i 1.  Still slows things down quite a bit and can't really do work while its mining. 

Try changing -H 2 as well with -i 1

[see360]

Try to rerun the autotune, I did that when the UTC moved to Nfactor 10 last week, it works okay.
You should also use -L2, you should gain some khash/s compared to the default setting (-L1)

Ran autotune for L=1, 2 & 3 for each card individually. It did find a faster config, thanks. Didn't try changing the -C -H or others, but here are the results.

780 Ti, Nfactor is 10 (N=2048)

cudaminer -d 0 -i 0 --algo=scrypt-jane:MRC -L 1 --benchmark
Maximum total warps (BxW): 321
T57x4  364793 hash/s
266 khash/s after running  <---------------- FASTEST

cudaminer -d 0 -i 0 --algo=scrypt-jane:MRC -L 2 --benchmark
Maximum total warps (BxW): 642
T15x23  276000 hash/s
221 khash/s after running

2nd attempt
T14x23  257600 hash/s
206 Khash/s after running

cudaminer -d 0 -i 0 --algo=scrypt-jane:MRC -L 3 --benchmark
Maximum total warps (BxW): 964
T29x23  194034.6 hash/s
166 khash/s after running

The best configs for the 660 Ti s end up with the same hash rate as before, but I've gone from 428 to about 485 khash/s.

[ivanlabrie]

Nice, I'll play some more with autotune...I always go manual. 
Takes too long otherwise :p

[djm34]

Try to rerun the autotune, I did that when the UTC moved to Nfactor 10 last week, it works okay.
You should also use -L2, you should gain some khash/s compared to the default setting (-L1)

Ran autotune for L=1, 2 & 3 for each card individually. It did find a faster config, thanks. Didn't try changing the -C -H or others, but here are the results.

780 Ti, Nfactor is 10 (N=2048)

cudaminer -d 0 -i 0 --algo=scrypt-jane:MRC -L 1 --benchmark
Maximum total warps (BxW): 321
T57x4  364793 hash/s
266 khash/s after running  <---------------- FASTEST

cudaminer -d 0 -i 0 --algo=scrypt-jane:MRC -L 2 --benchmark
Maximum total warps (BxW): 642
T15x23  276000 hash/s
221 khash/s after running

2nd attempt
T14x23  257600 hash/s
206 Khash/s after running

cudaminer -d 0 -i 0 --algo=scrypt-jane:MRC -L 3 --benchmark
Maximum total warps (BxW): 964
T29x23  194034.6 hash/s
166 khash/s after running

The best configs for the 660 Ti s end up with the same hash rate as before, but I've gone from 428 to about 485 khash/s.

Can you try this -L2 T30x16 on the gtx780ti. In principle the best configs are dividers of the cuda core number.

[see360]

Can you try this -L2 T30x16 on the gtx780ti. In principle the best configs are dividers of the cuda core number.

Thanks for the suggestion. I'm always looking to better understand the way CUDA sends threads in warps through the pipes.

-d 0 -scrypt-jane:MRC -L2 T30x16 --benchmark
225 khash/s, gets thermally throttled to 205.

I'm beginning to think my 780ti performance might be hampered by airflow between the card next to it. But it does give 650 khash/s with scrypt while all three are running (1205 total).

[ivanlabrie]

Can you try this -L2 T30x16 on the gtx780ti. In principle the best configs are dividers of the cuda core number.

Thanks for the suggestion. I'm always looking to better understand the way CUDA sends threads in warps through the pipes.

-d 0 -scrypt-jane:MRC -L2 T30x16 --benchmark
225 khash/s, gets thermally throttled to 205.

I'm beginning to think my 780ti performance might be hampered by airflow between the card next to it. But it does give 650 khash/s with scrypt while all three are running (1205 total).

You need water cooling
3 780ti's on air?

I only have one regular 780 and it does get hot mining scrypt (I don't like mining scrypt though) 716kh/s at 1267mhz core

[djm34]

I have the same problem since I added the R9290x, the gtx780ti gets hot and clocks down...
When I saw those saturday on the forum those riser with usb3.0 cable, I bought one with a cable 60cm so I can put the radeon far away from the rest of the computer

[ivanlabrie]

I have the same problem since I added the R9290x, the gtx780ti gets hot and clocks down...
When I saw those saturday on the forum those riser with usb3.0 cable, I bought one with a cable 60cm so I can put the radeon far away from the rest of the computer

They are pretty cool, and way tougher than the ide ribbon ones (broke a few of them)

[SUGO]

Sorry, but I  searched last pages but I did'nt find it:  how I can raise the clock gpu over the limit 1310 (100% in evga or afterburner)
I think about flash with kepler, but what parameter I should change?

[poohbah]

Sorry, but I  searched last pages but I did'nt find it:  how I can raise the clock gpu over the limit 1310 (100% in evga or afterburner)
I think about flash with kepler, but what parameter I should change?

This one?  http://cryptomining-blog.com/tag/gtx-750-ti-cudaminer/

[ivanlabrie]

Sorry, but I  searched last pages but I did'nt find it:  how I can raise the clock gpu over the limit 1310 (100% in evga or afterburner)
I think about flash with kepler, but what parameter I should change?

Afterburner sucks for nvidia, use EVGA Precision X.

[69charger]

Well, time for a more capable MoBo 

http://www.newegg.com/Product/Product.aspx?Item=N82E16813131819

Asus P8Z77-V Pro does not like to play well with 3 graphics cards...

The third slot is bandwidth limited and in order to use it at 4X the board disables a bunch of other onboard stuff including sound and the other pci-e1x slots.

Ghey....

[cvax]

Hi guys, I am trying to mine Yac with my GTX 590, but it isn't liking my configuration. It works with autotune, but when I use the exact same -l setting that was found in autotune it just completely blows up as in this screenshot. Any ideas? Thanks.


Here is my command line.
cudaminer.exe --algo=scrypt-jane -H 1 -i 0 -l f8x8 -C 2 -o stratum+tcp://yac.coinmine.pl:9088

[sub-80]

Gonna try this cool winpe 64 with the cudaminer 64bit binaries...

http://windowsmatters.com/

thoughts?

That your awesome

[sub-80]

Hi guys, I am trying to mine Yac with my GTX 590, but it isn't liking my configuration. It works with autotune, but when I use the exact same -l setting that was found in autotune it just completely blows up as in this screenshot. Any ideas? Thanks.


Here is my command line.
cudaminer.exe --algo=scrypt-jane -H 1 -i 0 -l f8x8 -C 2 -o stratum+tcp://yac.coinmine.pl:9088

--algo=scrypt-jane:YAC

[DeltaQuebec]

May need help, or an answer with 2014.2.18 cudaminer mining CACHE coin.

I believe now Cache is now N-Factor 2048 but im really shocked at the slow hash rate I am getting
59-60 kh/s .. usually these N-scrypt coins I get near 100 kh/s, but this is 1/2 of that, is this normal?

Bat: cudaminer.exe  --algo=scrypt-jane:CACH -H 1 -i 0 -C 0 -m 1 -L 1 -l k4x28 -o
stratum+tcp://east1.us.stratum.dedicatedpool.com:3364......

Card is a GTX 650 Ti 8GB system ram ( 1GB on card )



I just want to be sure im doing things right.. but pretty shocked at the slow rate.

Thanks in advance!