this image suggests to use http:// and not stratum+tcp://

http://imgur.com/r/all/cesAJhA

Maybe the stratum would require a different port number?

Okay, i made some improvements to the Titan kernel, which brought my khash/s up from 55 kHash (achieved with the Kepler Kernel) to 62 kHash on the GT 640 (GK208 chip, Compute 3.5). Maybe you want to try this binary on the GTX 780? Let me know how I should send it to you.

Christian

Must let it autotune. then use -l once it find a valid configuration..Plenty of answers to this..means you've chosen an invalid configuration.

make sure drivers are up to date.

there new switches like -K and -F Christian posted them a couple pages back. I would suggest you try them. You could also disable the 9400gt and try just mining with the gtx295, then do the same with the other one. to try to determine if one is working or not

I'm trying to run cudaminer for the first time on my GTX 480 SLI cards.  This is Windows 8 (not 8.1 to my knowledge) with latest beta driver (331.65) from Geforce experience.  When I autotune it crashes after picking a kernel and I'm not sure how to get useful info.  I do have some of the CUDA SDK installed on my computer, but haven't touched it recently.  Was years ago I last programmed something CUDA, and installed it more recently on a whim.  The last few lines of a failing run with debug turned on is below...

>cudaminer -D -H 1 -i 1,0 -o stratum+tcp://<url> -O <user>:<pass>

...
[2013-11-06 20:38:56] 353:     |     |     |     |     |     |     |     |     |     |     |     |
   |     |     |      kH/s
[2013-11-06 20:38:56] 354:     |     |     |     |     |     |     |     |     |     |     |     |
   |     |     |      kH/s
[2013-11-06 20:38:56] GPU #0:   71.63 khash/s with configuration F7x8
[2013-11-06 20:38:56] GPU #0: using launch configuration F7x8


I tried the idea of limiting which kernels to use.  When I picked Fermi (which my card is) it crashed again.  When I picked Kepler I got farther, but this still seems wrong not least of which because this is not a Kepler based card and the errors from card 1.

>cudaminer -H 1 -i 1,0 -l K -o stratum+tcp://<url> -O <user>:<pass>
           *** CudaMiner for nVidia GPUs by Christian Buchner ***
                     This is version 2013-11-01 (alpha)
        based on pooler-cpuminer 2.3.2 (c) 2010 Jeff Garzik, 2012 pooler
               Cuda additions Copyright 2013 Christian Buchner
           My donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm

[2013-11-07 18:18:25] 2 miner threads started, using 'scrypt' algorithm.
[2013-11-07 18:18:25] Starting Stratum on stratum+tcp://<url>
[2013-11-07 18:18:27] GPU #1: GeForce GTX 480 with compute capability 2.0
[2013-11-07 18:18:27] GPU #1: interactive: 0, tex-cache: 0 , single-alloc: 0
[2013-11-07 18:18:27] GPU #0: GeForce GTX 480 with compute capability 2.0
[2013-11-07 18:18:27] GPU #0: interactive: 1, tex-cache: 0 , single-alloc: 0
[2013-11-07 18:18:27] GPU #1: Performing auto-tuning (Patience...)
[2013-11-07 18:18:27] GPU #0: Given launch config 'K' does not validate.
[2013-11-07 18:18:27] GPU #0: Performing auto-tuning (Patience...)
[2013-11-07 18:19:07] GPU #0:  201.97 khash/s with configuration K15x16
[2013-11-07 18:19:07] GPU #0: using launch configuration K15x16
[2013-11-07 18:19:07] GPU #0: GeForce GTX 480, 7680 hashes, 0.19 khash/s
[2013-11-07 18:19:07] GPU #0: GeForce GTX 480, 15360 hashes, 128.99 khash/s
[2013-11-07 18:19:36] GPU #1:  213.19 khash/s with configuration F30x8
[2013-11-07 18:19:36] GPU #1: using launch configuration F30x8
[2013-11-07 18:19:36] GPU #1: GeForce GTX 480, 7680 hashes, 0.11 khash/s
[2013-11-07 18:19:36] GPU #1: GeForce GTX 480, 7680 hashes, 105.13 khash/s
[2013-11-07 18:19:36] GPU #1: GeForce GTX 480 result does not validate on CPU!
[2013-11-07 18:19:40] GPU #0: GeForce GTX 480, 6174720 hashes, 190.56 khash/s
[2013-11-07 18:19:40] accepted: 1/1 (100.00%), 295.70 khash/s (yay!!!)
[2013-11-07 18:19:43] GPU #0: GeForce GTX 480, 698880 hashes, 184.91 khash/s
[2013-11-07 18:19:43] accepted: 2/2 (100.00%), 290.04 khash/s (yay!!!)
[2013-11-07 18:19:46] GPU #1: GeForce GTX 480, 2181120 hashes, 209.68 khash/s
[2013-11-07 18:19:47] accepted: 3/3 (100.00%), 394.59 khash/s (yay!!!)
Ctrl-C
[2013-11-07 18:19:49] workio thread dead, waiting for workers...
[2013-11-07 18:19:49] worker threads all shut down, exiting.


Thanks for any help in advance.

----------------------------------

Just realized SLI was turned off in the driver.  I'll try with it on in a bit...

I scrapped both my GTX 260 (SLI config) because I was getting only 40 kHash/s from each one under Windows 7.
They might have worked better on Linux or Windows XP (due to a different driver model used there).

I also scrapped a GTX 460 (too old, 96 kHash/s).

I got myself instead:
a GTX 560 Ti 448 core edition (used), a GTX 560 Ti (new) and a GTX 660 Ti (new) and a GT 640 (the new model with GK 208 chip).

Now the machine can do 600 kHash @ 800 Watts. And it's good for driving MANY monitors and for gaming as well And it isn't actually profitable, considering the local electricity costs and the increased mining difficulty. But I am doing this for fun.

Christian

ever tried running two separate instances of cudaminer? one with -d 0 and one with -d 1 maybe?

do tools like CUDA-z and GPU-z show both chips separately?

Christian

When I start the program it just shuts down immediately.

I navigated via cmd and when running it, it just tells me about the program, the developed and how to donate - it never starts

halp!

I have both instances set to K16x16 and they both maintain around 145 khash/s independently, but one always seems to overtake the other in accepted blocks (by a decent amount). What I am doing now is splitting my mining up between FTC and LTC. One GPU mines litecoin while the other is mining feathercoin. It's actually kind of nice, but I would still like to find a way to put them together within one instance to ramp up the output on just one.