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

[MexiMelt]

Increase -L to 1024 and doubling works :-)

I did, doesn't work for me.

[bigjme]

On a 780 or something lower?

[MexiMelt]

On a 780 or something lower?

It's a GTX 670. has 1344 cuda cores. 1344x32 works at -L 512. No matter what I specify for -L, doubling doesn't work. Out of memory every time.

[Legionnairez]

I received the halving in hashrate again with the newest build, I think I found it... trying to replicate it, will edit here soon.
E: see this post below.

[cbuchner1]

780ti SC non/ACX - +51 offset - 220,000 KHash.

Wow! Much hash. I get only 180 per 780 Ti on Linux, and I lack all overclocking options
except for BIOS mods

Christian

[bathrobehero]

I received the halving in hashrate again with the newest build, I think I found it... trying to replicate it, will edit here soon.

You're driver crashed.

[cbuchner1]

It's a GTX 670. has 1344 cuda cores. 1344x32 works at -L 512. No matter what I specify for -L, doubling doesn't work. Out of memory every time.

but can't you just ignore the out of memory? it's for scrypt related buffers mostly, which keccak doesn't need.

maybe -L 1024 -m 1 does the trick?

[cbuchner1]

Nothing beats Y1024x32 on my 660.

If I would have more time I'd figure out to setup autotune so that it starts testing from like Y700x

autotune would measure scrypt, so that's no use....

[Legionnairez]

I received the halving in hashrate again with the newest build, I think I found it... trying to replicate it, will edit here soon.

You're driver crashed.

I know, I'm trying to replicate what causes it so it can be fixed - unless you know what causes this

[bathrobehero]

I received the halving in hashrate again with the newest build, I think I found it... trying to replicate it, will edit here soon.

You're driver crashed.

I know, I'm trying to replicate what causes it so it can be fixed - unless you know what causes this

Well, pretty much everything. High overclock, unstable launch or even closing cudaMiner does that for me almost every time (CTRL+C or just closing the window is fine, but taskkill or --time-limit crashes very often with keccak).

Edit: Awww, would you look at that, 100%

[DirtyWilly]

780ti SC non/ACX - +51 offset - 220,000 KHash.

Wow! Much hash. I get only 180 per 780 Ti on Linux, and I lack all overclocking options
except for BIOS mods

Christian

Zero booos after the initial start up.  Still running strong at 220,000 khash after 25 minutes.

https://i.imgur.com/ayqSQxm.png

I'm telling you open Chrome or something that makes the GPU stepping increase?  Also run maxcoind with two less cores in affinity.  Yields extra hash for unknown reason?

My 680 GTX SC is solid 127,000 khash, same settings.  99%.

Thanks Christian!

[old_pioner]

For the kernel launch configuration: Take the total number of CUDA cores on your card, multiply by 2, and append x32 for the K kernel and x16 for the F kernel.

Yes, terrible autotuning

Code:

[2014-02-10 01:30:44] Starting Stratum on stratum+tcp://maxpool.1gh.com:17333
[2014-02-10 01:30:44] 1 miner threads started, using 'keccak' algorithm.
[2014-02-10 01:30:45] GPU #0: GeForce GTX 650 Ti with compute capability 3.0
[2014-02-10 01:30:45] GPU #0: interactive: 0, tex-cache: 1D, single-alloc: 1
[2014-02-10 01:30:45] GPU #0: 8 hashes / 1.0 MB per warp.
[2014-02-10 01:30:45] GPU #0: Performing auto-tuning (Patience...)
[2014-02-10 01:30:45] GPU #0: maximum total warps (BxW): 856
[2014-02-10 01:33:25] Stratum detected new block
[2014-02-10 01:34:19] GPU #0: 7632054528.00 hash/s with configuration k61x14
[2014-02-10 01:34:19] GPU #0: using launch configuration k61x14
[2014-02-10 01:34:20] GPU #0: GeForce GTX 650 Ti, 0.00 khash/s
[2014-02-10 01:34:20] GPU #0: GeForce GTX 650 Ti, 3729 khash/s
[2014-02-10 01:34:37] Stratum detected new block
[2014-02-10 01:34:37] GPU #0: GeForce GTX 650 Ti, 5553 khash/s
[2014-02-10 01:35:15] GPU #0: GeForce GTX 650 Ti result for nonce $0d3a4d37 does
 not validate on CPU!
[2014-02-10 01:35:15] GPU #0: GeForce GTX 650 Ti result for nonce $0d3a4d37 does
 not validate on CPU!
[2014-02-10 01:35:15] GPU #0: GeForce GTX 650 Ti result for nonce $0d3a4d37 does
 not validate on CPU!
[2014-02-10 01:35:15] GPU #0: GeForce GTX 650 Ti, 5850 khash/s

-l F192x16 --algo=keccak -m 1 -i 0 -H 2 -C 2

Code:

[2014-02-10 02:08:41] GPU #1: GeForce GT 440 with compute capability 2.1
[2014-02-10 02:08:41] GPU #1: interactive: 0, tex-cache: 2D, single-alloc: 1
[2014-02-10 02:08:41] GPU #1: 32 hashes / 0.0 MB per warp.
[2014-02-10 02:08:41] GPU #1: using launch configuration F192x16
[2014-02-10 02:08:41] GPU #1: GeForce GT 440, 15379 khash/s
[2014-02-10 02:08:43] Stratum detected new block
[2014-02-10 02:08:43] GPU #1: GeForce GT 440, 16178 khash/s

-l Y1536x32 --algo=keccak -m 1 -i 0 -H 2 -C 1

Code:

[2014-02-10 02:00:52] GPU #0: GeForce GTX 650 Ti with compute capability 3.0
[2014-02-10 02:00:52] GPU #0: interactive: 0, tex-cache: 1D, single-alloc: 1
[2014-02-10 02:00:52] GPU #0: 32 hashes / 0.0 MB per warp.
[2014-02-10 02:00:52] GPU #0: Launch config 'Y1536x32' requires too much memory!

[2014-02-10 02:00:52] GPU #0: using launch configuration K1536x32
[2014-02-10 02:00:52] GPU #0: GeForce GTX 650 Ti, 40265 khash/s
[2014-02-10 02:00:55] GPU #0: GeForce GTX 650 Ti, 59769 khash/s
[2014-02-10 02:00:55] accepted: 1/1 (100.00%), 59769 khash/s (yay!!!)
[2014-02-10 02:01:19] GPU #0: GeForce GTX 650 Ti, 60802 khash/s

There is no kernel for mining Protoshares?

[Legionnairez]

I received the halving in hashrate again with the newest build, I think I found it... trying to replicate it, will edit here soon.

You're driver crashed.

I know, I'm trying to replicate what causes it so it can be fixed - unless you know what causes this

Well, pretty much everything. High overclock, unstable launch or even closing cudaMiner does that for me almost every time.

I'm running stable on +75/+75 overclocks, but the following seems to cause the driver crash/ hashrate halving:
If I try to mine YAC with the following .bat:
cudaminer.exe --algo=scrypt-jane -i 1 -H 1 -l K8x32
I receive the following error:

On a Zotac GTX 680 4GB.

Haven't had the time to even try to solve this, besides bed time - Once again a tired monday after staying up solving mining stuff ^^

[djm34]

what exactly are protoshare ?

[Mirkic7]

I'm telling you open Chrome or something that makes the GPU stepping increase?  Also run maxcoind with two less cores in affinity.  Yields extra hash for unknown reason?

Just put in maxcoin.conf option:
gen=0

This will disable CPU mining, and keep low CPU usage on it.

->> http://maxpool.sytes.net <<-
Need more workers!
0% fee, stratum
cgminer/cudamier/cpuminer supported
Correct HashRates
Based on MPOS

1st block finder gets 1 MAX bonus!

[bathrobehero]

->> http://maxpool.sytes.net <<-
Need more workers!
0% fee, stratum
cgminer/cudamier/cpuminer supported
Correct HashRates
Based on MPOS

1st block finder gets 1 MAX bonus!

Might as well go for that 1 MAX bonus ^^

[DirtyWilly]

Christian, what is your BTC address?

[Legionnairez]

Christian, what is your BTC address?

From the readme:
If you find this tool useful and like to support its continued
          development, then consider a donation.

  LTC donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm
  BTC donation address: 16hJF5mceSojnTD3ZTUDqdRhDyPJzoRakM
  YAC donation address: Y87sptDEcpLkLeAuex6qZioDbvy1qXZEj4
  VTC donation address: VrjeFzMgvteCGarLw85KivBzmsiH9fqp4a
  MAX donation address: mHrhQP9EFArechWxTFJ97s9D3jvcCvEEnt

[DirtyWilly]

Christian, what is your BTC address?

From the readme:
If you find this tool useful and like to support its continued
          development, then consider a donation.

  LTC donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm
  BTC donation address: 16hJF5mceSojnTD3ZTUDqdRhDyPJzoRakM
  YAC donation address: Y87sptDEcpLkLeAuex6qZioDbvy1qXZEj4
  VTC donation address: VrjeFzMgvteCGarLw85KivBzmsiH9fqp4a
  MAX donation address: mHrhQP9EFArechWxTFJ97s9D3jvcCvEEnt

Just making sure this is up to date.

[taltamir]

Do I need to download Cuda or standard drivers are enough?

standard drivers run cudaminer just fine