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

[djm34]

may-be the twin frozr, but with that difference I doubt it

djm34 you could try xintensity 105 or 108. My 290 get 8.3Mhs from this.

yes it is a bit better at 105, I get an average at 8.7Mhash/s.
At 108, it is a bit all over the place with up to 14 but it doesn't stay there and I get lot of rejected

[NeuroticFish]

Consider asking the botnet guy or the big farm guy and look at Christian numbers

Anybody from top10 makes a real difference.
And when smaller pools get enough MH/s the small guys will follow.

The botnet? I asked Christian because he seems a smart and reasonable guy. So there's a chance. I don't know the botnet guys and I doubt they'd listen anyway.

[djm34]

Consider asking the botnet guy or the big farm guy and look at Christian numbers

Anybody from top10 makes a real difference.
And when smaller pools get enough MH/s the small guys will follow.

The botnet? I asked Christian because he seems a smart and reasonable guy. So there's a chance. I don't know the botnet guys and I doubt they'd listen anyway.

Actually he might be the only one who can really move as the cgminer is for 1gh only

[ivanlabrie]

Damn, let's hope a gpu miner comes out fast for the other pools...otherwise this could kill the coin before I cash out xD

[NeuroticFish]

Actually he might be the only one who can really move as the cgminer is for 1gh only

Oh, this may be a problem indeed, though nonce pool works with the 1gh miner....

Please ignore this post, I've misread yours

[reorder]

So anybody knows how to contact nonce-pool ops so they could look into incompatibility with cgminer? (as off the topic as it is). The incompatibility is not intentional in any case, just there is some subtle implementation difference.

[cbuchner1]

Actually he might be the only one who can really move as the cgminer is for 1gh only

the miner also works on http://hvc.nonce-pool.com/  - this was tested.

The coin however has lost so much value that I don't see any long term profitability in it anymore.
My 10000 GPU mined HVC are worth like...0.2 BTC. This is a joke (well compared to Maxcoin,
that is...)

And why should anyone trust this coin now that their sacred mantra of being "GPU and ASIC
resistant" has been proven wrong. You really need to design a coin with such aspirations
based on algorithms that are *known* to be hard to parallelize. You don't just take SHA256
and add a bit of sparkly fairy dust in hopes of making it un-parallelizable. And calling the
algorithm HEFTY1 doesn't change that either.



Christian

[cvax]

Look,I bought a video card with wrong GPU  

Wait... what?? Which brand was this?

[ken-ray]

again, two short questions:

1. why can't i clone from github as suggested in the start post? I always get a 403 :/
2. what linux do you guys suggest for a 5 GPU rig? I'd like to use debian if possible. Is everything available for that distro?

I use debian sid and it works ok. Drivers and cuda in repository work with GTX 750 Ti. I haven't tried 5 gpus.

[old_pioner]

Wait... what?? Which brand was this?

http://www.palit.biz/palit/vgapro.php?id=2253
They wrote CUDA Cores    640, but make 960 for me only maybe  

[mhps]

I am mining YACs with C2750's. Cudaminer dies randomly with lots of #30 errors. Is it possible to let the miner sleep for 10 seconds and re-start from the beginning again when such error happens? Or at least exit with an error code so that a script can be written to restart the miner? Thanks.

The miner is the 2014-02-28 one, running on win8, with "-m 1 -i 1 -H 1 -C 1". The errors are

Code:

[2014-03-17 xx:xx:xx] GPU #0: cudaError 30 (unknown error) calling 'cudaMemcpyAs
ync(context_idata[stream][thr_id], X, mem_size, cudaMemcpyHostToDevice, context_
streams[stream][thr_id])' (D:/Christian/Documents/Visual Studio 2010/Projects/Cu
daMiner/salsa_kernel.cu line 938)

ed: error msg and command line settings

[djm34]

Wait... what?? Which brand was this?

http://www.palit.biz/palit/vgapro.php?id=2253
They wrote CUDA Cores    640, but make 960 for me only maybe 

or gpu-z is not up to date...

[Bombadil]

I am mining YACs with C2750's. Cudaminer dies randomly with lots of #30 errors. Is it possible to let the miner sleep for 10 seconds and re-start from the beginning again when such error happens? Or at least exit with an error code so that a script can be written to restart the miner? Thanks.

I'm using this for my bat for my gtx760.

Code:

:start
cudaminer.exe  -o STRATUM -O LOGIN -H 1 -i 0 -C 1 -l K6x32 --time-limit 3600
goto start

It kills the cudaminer process after an hour, and restarts the loop. I was getting many different errors with some stratums after 2 hours or so, no problems ever since I've been using this loop.

[mhps]

I'm using this for my bat for my gtx760.

Code:

:start
cudaminer.exe  -o STRATUM -O LOGIN -H 1 -i 0 -C 1 -l K6x32 --time-limit 3600
goto start

It kills the cudaminer process after an hour, and restarts the loop. I was getting many different errors with some stratums after 2 hours or so, no problems ever since I've been using this loop.

Thanks Bombadil. I will try your script with "nvidia-smi >> log" to record how the card has been working.

[Lacan82]

I'm almost considering making maxwell rig soon. I've also noticed coinwars.com has added. Quark Algo to their calculator. Maybe we'll see other soon

[cvax]

Actually he might be the only one who can really move as the cgminer is for 1gh only

the miner also works on http://hvc.nonce-pool.com/  - this was tested.

The coin however has lost so much value that I don't see any long term profitability in it anymore.
My 10000 GPU mined HVC are worth like...0.2 BTC. This is a joke (well compared to Maxcoin,
that is...)

And why should anyone trust this coin now that their sacred mantra of being "GPU and ASIC
resistant" has been proven wrong. You really need to design a coin with such aspirations
based on algorithms that are *known* to be hard to parallelize. You don't just take SHA256
and add a bit of sparkly fairy dust in hopes of making it un-parallelizable. And calling the
algorithm HEFTY1 doesn't change that either.



Christian

Looks like your price has recovered a bit. Hope you cash out alright.

[jack80]

Wait... what?? Which brand was this?

http://www.palit.biz/palit/vgapro.php?id=2253
They wrote CUDA Cores    640, but make 960 for me only maybe  

OK! its a bomb news!

download gpuz 0.7.7 http://www.techpowerup.com/downloads/2340/techpowerup-gpu-z-v0-7-7/

and add more pictures

It is not just a matter of gpu z but also the miner to mine more than 400 kh / s

can you give us the exact serial number of the card?

What do I think?

That perhaps all current Nvidia 750ti are born with 960 Cuda Core and suffice to update the BIOS to unlock them!

.

Look,I bought a video card with wrong GPU  

.

[liomojo1]

Look I have 7 of the exact the same card and it doesn`t have 960 cuda and makes 300khs overclocked.

[MarcusDe]

is this real?

For the impatient, here is HVC cgminer version that works on 1gh: https://github.com/reorder/cgminer_skein

It currently only works with GCN cards (or so I believe). No binaries are available as it is a WIP, and I will not provide support for this version.

Please configure with './autogen.sh --enable-heavy --enable-opencl' and run with '--heavy --vote 1024' (or your preferred vote). Best intensity setting seems to be 3 for all cards, but you may have luck with higher values.

I'm sorry to say but Christian, this time AMD people again have more advantage than us. I confirm that this miner works with AMD cards.
Expect the prices drop of HVC like MAX again.

confirm, it works, 5.5Mhash/s with R9290x (and I think it is already too late... everybody seems to be getting many mhash on 1gh... )
want a real cpu only coin (due to some misconfig, miners work only from the wallet  ): rhinohorn

--------------------------------------------
 GPU 1:  45.0C 2986RPM | 6.416M/6.402Mh/s |
 GPU 2:  48.0C 3062RPM | 6.407M/6.402Mh/s |
 GPU 3:  47.0C 3074RPM | 6.416M/6.402Mh/s |
 GPU 4:  43.0C 2940RPM | 6.409M/6.401Mh/s |
--------------------------------------------

Thats my 4x 280X

[ltcnim]

there's 2 750TI gpus, and i always thought, that once a share is found, the full hashing power (added from all cards) is shown. This would mean, he only gets 200khash per card.