Bitcoin Forum
April 19, 2024, 09:36:52 PM *
News: Latest Bitcoin Core release: 26.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 [34] 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 ... 1135 »
  Print  
Author Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX]  (Read 3426867 times)
shatter
Newbie
*
Offline Offline

Activity: 28
Merit: 0


View Profile
May 01, 2013, 10:55:08 AM
 #661

-l S49x4 got me to 170 khash+

GTX 670
Activity + Trust + Earned Merit == The Most Recognized Users on Bitcointalk
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
cbuchner1 (OP)
Hero Member
*****
Offline Offline

Activity: 756
Merit: 502


View Profile
May 01, 2013, 10:59:50 AM
 #662

-l S49x4 got me to 170 khash+

a few postings up, someone said he would get 185 kHash on GTX 670 with this.

-l S56x4 -C 1
yakka34
Newbie
*
Offline Offline

Activity: 11
Merit: 0



View Profile
May 01, 2013, 11:21:18 AM
 #663

-l S49x4 got me to 170 khash+

a few postings up, someone said he would get 185 kHash on GTX 670 with this.

-l S56x4 -C 1

I only got 178kh/s with those.Mine flags  -C 2 -i 0 -l S112x2 gives me 182-185kh/s. I've also added these flags to comparisons spreadsheet.
shatter
Newbie
*
Offline Offline

Activity: 28
Merit: 0


View Profile
May 01, 2013, 11:27:37 AM
 #664

-l S49x4 got me to 170 khash+

a few postings up, someone said he would get 185 kHash on GTX 670 with this.

-l S56x4 -C 1

I only got 178kh/s with those.Mine flags  -C 2 -i 0 -l S112x2 gives me 182-185kh/s. I've also added these flags to comparisons spreadsheet.

Nice one! Thanks coming close to 180 now
termhn
Full Member
***
Offline Offline

Activity: 126
Merit: 100


View Profile
May 01, 2013, 02:13:50 PM
 #665

-l S49x4 got me to 170 khash+

a few postings up, someone said he would get 185 kHash on GTX 670 with this.

-l S56x4 -C 1

I only got 178kh/s with those.Mine flags  -C 2 -i 0 -l S112x2 gives me 182-185kh/s. I've also added these flags to comparisons spreadsheet.
Oh sweet I'm at 190 now Smiley
Mikanoshi
Sr. Member
****
Offline Offline

Activity: 798
Merit: 252


Insane In The Blockchain ⚠


View Profile WWW
May 01, 2013, 02:33:29 PM
 #666

-C 2 -i 0
Autotuned S111x2 = 195
S112x2 gives 199 @ 1201 MHz

airforcc
Newbie
*
Offline Offline

Activity: 49
Merit: 0


View Profile
May 01, 2013, 02:53:54 PM
 #667

Hi cbuchner1,

Thanks for your great work.

Anyway in China it is very difficult to fetch software from Dropbox /mega.co.nz.

Would you please put the md5sum / sha256sum to this thread so others can
download cudaminer from mirror and verify it here?

Thanks again.
tacotime
Legendary
*
Offline Offline

Activity: 1484
Merit: 1005



View Profile
May 01, 2013, 06:28:44 PM
 #668

Hey, I need something for you to properly implement this in GUIminer, which is to allow a flag that updates the hash rate piped to the console at a given rate, eg 1 line containing hash rate and hashes per second.  Then it should be ready to go.

Aggregated over all GPUs? Isn't the term hash rate synonymous to hashes per second?

I will see what I can do.

Okay, so this is what I exactly need if you have time and the inclination to help:

- argument flag that enables a "quiet" mode
- "quiet" mode entails starting the miner, then reporting the hash rate/hashes every second as the program normally does
- accepted/rejected messages are reported after a share is submitted
- everything is passed directly to stdout rather than using applog (since it mangles the output for python).  not sure if it matters what function (cout or printf)

This is almost exactly like -T mode for cgminer.  The weird thing is that the miner looks fine in console but subprocess can't see any of the data the program is outputting -- like it doesn't use stdout/stderr or something.

If it still fails to output correctly I'll try to write a hook program in C++/python and play around with it...

Code:
XMR: 44GBHzv6ZyQdJkjqZje6KLZ3xSyN1hBSFAnLP6EAqJtCRVzMzZmeXTC2AHKDS9aEDTRKmo6a6o9r9j86pYfhCWDkKjbtcns
Amph
Legendary
*
Offline Offline

Activity: 3206
Merit: 1069



View Profile
May 01, 2013, 08:27:18 PM
 #669

it's strange, the program start only with bitcoin, but not with litecoin...
with litecoin it give me an error
Lacan82
Sr. Member
****
Offline Offline

Activity: 247
Merit: 250


View Profile
May 01, 2013, 08:54:03 PM
 #670

it's strange, the program start only with bitcoin, but not with litecoin...
with litecoin it give me an error

What command are you trying to execute? are drivers up to date?

termhn
Full Member
***
Offline Offline

Activity: 126
Merit: 100


View Profile
May 01, 2013, 09:29:11 PM
 #671

it's strange, the program start only with bitcoin, but not with litecoin...
with litecoin it give me an error
This is meant only to work with litecoin not bitcoin
SubNoize
Newbie
*
Offline Offline

Activity: 47
Merit: 0


View Profile
May 01, 2013, 11:56:15 PM
 #672

This is written in the Readme, hope it helps:

Currently there is just one prefix, which is "S". Later releases may
see the introduction of more kernel variants with using other letters.

Examples:

e.g. S27x3 is a launch configuration that works well on GTX 260
      28x4 is a launch configuration that works on Geforce GTX 460
     290x2 is a launch configuration that works on Geforce GTX 660Ti

You should wait through autotune to see what kernel is found best for
your current hardware configuration.

The choice between Non-Titan and Titan CUDA kernels is automatically
made based on your device's compute capability. Titans cost around
a thousand dollars, so you probably don't have one.


Prefix  | Non-Titan          | Titan
-------------------------------------------------------
 <none> | low shared memory  | default kernel
        | optimized kernel   | with funnel shifter
        |                    |
   S    | spinlock kernel    | spinlock kernel
        | for Kepler GPUs    | with funnel shifter

Can anyone explain like i'm 5? If my card used to autotune to the default kernal and is now auto tuning to the spinlock kernel will it actually make a difference?
termhn
Full Member
***
Offline Offline

Activity: 126
Merit: 100


View Profile
May 02, 2013, 12:01:03 AM
 #673

If you have a Kepler (600 series) card it will. Otherwise it won't.
SubNoize
Newbie
*
Offline Offline

Activity: 47
Merit: 0


View Profile
May 02, 2013, 01:28:33 AM
 #674

i'm on a 580 and auto tune makes it use S launch config. Is this a bug?
termhn
Full Member
***
Offline Offline

Activity: 126
Merit: 100


View Profile
May 02, 2013, 01:44:48 AM
 #675

i'm on a 580 and auto tune makes it use S launch config. Is this a bug?
No clue
Stoneysilence
Member
**
Offline Offline

Activity: 104
Merit: 10


View Profile
May 02, 2013, 02:53:39 AM
Last edit: May 02, 2013, 03:45:20 AM by Stoneysilence
 #676

I upgraded to the latest version (2nd upload for the day of 4-30-13) and a few seconds after launching I got an error saying:
[2013-05-01 19:49:58] GPU #0: GeForce GTX 460 result does not validate on CPU!
Code:
C:\Mining\cudaminer-2013-04-30>cudaminer.exe -d 0 -i 0 -m 0 -C 0 -o http://invas
ionnetwork.com:8336 -O username:pass
           *** CudaMiner for nVidia GPUs by Christian Buchner ***
                     This is version 2013-04-30 (alpha)
        based on pooler-cpuminer 2.2.3 (c) 2010 Jeff Garzik, 2012 pooler
               Cuda additions Copyright 2013 Christian Buchner
           My donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm

[2013-05-01 19:48:12] 1 miner threads started, using 'scrypt' algorithm.
[2013-05-01 19:48:13] Long-polling activated for http://invasionnetwork.com:8336
/lp
[2013-05-01 19:48:13] GPU #0: GeForce GTX 460 with compute capability 2.1
[2013-05-01 19:48:13] GPU #0: interactive: 0, tex-cache: 0 , single-alloc: 0
[2013-05-01 19:48:13] GPU #0: Performing auto-tuning (Patience...)
[2013-05-01 19:49:29] GPU #0:  112.00 khash/s with configuration  14x8
[2013-05-01 19:49:29] GPU #0: using launch configuration  14x8
[2013-05-01 19:49:29] GPU #0: GeForce GTX 460, 7168 hashes, 0.09 khash/s
[2013-05-01 19:49:30] GPU #0: GeForce GTX 460, 7168 hashes, 91.90 khash/s
[2013-05-01 19:49:45] GPU #0: GeForce GTX 460, 1756160 hashes, 112.29 khash/s
[2013-05-01 19:49:45] accepted: 1/1 (100.00%), 112.29 khash/s (yay!!!)
[2013-05-01 19:49:58] GPU #0: GeForce GTX 460, 1394176 hashes, 112.52 khash/s
[2013-05-01 19:49:58] GPU #0: GeForce GTX 460 result does not validate on CPU!

I had this happen before on my Laptop once before where it constantly gave me that error.  I had to stop the program and restart it and then that error message went away.

Any idea's to what is causing it?
SubNoize
Newbie
*
Offline Offline

Activity: 47
Merit: 0


View Profile
May 02, 2013, 03:13:25 AM
 #677

No idea but there is 2 newer versions. Try upgrading to those.
Stoneysilence
Member
**
Offline Offline

Activity: 104
Merit: 10


View Profile
May 02, 2013, 03:44:50 AM
 #678

No idea but there is 2 newer versions. Try upgrading to those.

I just downloaded this one like 10 minutes before I posted from Page 1.  There is no newer version.

It was a typo on my part where I typed 4-30-22 meant 4-30-13.
Stoneysilence
Member
**
Offline Offline

Activity: 104
Merit: 10


View Profile
May 02, 2013, 04:43:30 AM
Last edit: May 02, 2013, 05:18:40 AM by Stoneysilence
 #679

I don't know if this is helpful to anybody but this is my Debug info for a GTX 460 to see what seems to be ideal.  Running the 4-30-13 2nd upload. I have noticed it seems to vary on what it thinks is the best.  Before upgrading to the 2nd upload I know it was running 14x8, now it seems to prefer 14x7.
Code:
C:\Mining\cudaminer-2013-04-30>cudaminer.exe -d 0 -D -i 1 -C 0 -o http://lo
calhost:8332 -O user:pass
           *** CudaMiner for nVidia GPUs by Christian Buchner ***
                     This is version 2013-04-30 (alpha)
        based on pooler-cpuminer 2.2.3 (c) 2010 Jeff Garzik, 2012 pooler
               Cuda additions Copyright 2013 Christian Buchner
           My donation address: LKS1WDKGED647msBQfLBHV3Ls8sveGncnm

[2013-05-01 21:35:06] 1 miner threads started, using 'scrypt' algorithm.
[2013-05-01 21:35:06] Long-polling activated for http://localhost:8332/lp
[2013-05-01 21:35:06] DEBUG: got new work in 15 ms
[2013-05-01 21:35:06] GPU #0: GeForce GTX 460 with compute capability 2.1
[2013-05-01 21:35:06] GPU #0: interactive: 1, tex-cache: 0 , single-alloc: 1
[2013-05-01 21:35:08] GPU #0: Performing auto-tuning (Patience...)
[2013-05-01 21:35:09]        x1    x2    x3    x4    x5    x6    x7    x8
[2013-05-01 21:35:09]   1:  1.5|  3.1|  4.6|  6.2|  6.2|  9.2| 10.8| 12.3 kh/s
[2013-05-01 21:35:09]   2:  3.1|  6.1|  9.2| 12.3| 15.4| 18.5| 21.5| 19.7 kh/s
[2013-05-01 21:35:10]   3:  4.6| 12.3| 13.8| 18.5| 23.1| 27.7| 25.8| 36.9 kh/s
[2013-05-01 21:35:10]   4:  6.2| 14.0| 18.5| 24.6| 30.8| 29.5| 43.1| 49.2 kh/s
[2013-05-01 21:35:11]   5:  7.7| 15.4| 23.1| 30.8| 38.5| 46.2| 43.1| 61.5 kh/s
[2013-05-01 21:35:11]   6:  7.4| 18.5| 27.7| 36.9| 46.2| 44.3| 64.6| 73.8 kh/s
[2013-05-01 21:35:12]   7: 10.8| 21.5| 32.3| 34.5| 53.8| 64.6| 60.3| 86.2 kh/s
[2013-05-01 21:35:12]   8: 12.3| 24.6| 36.9| 49.2| 49.2| 59.1| 57.4| 65.6 kh/s
[2013-05-01 21:35:13]   9: 18.5| 27.7| 33.2| 55.4| 55.4| 66.5| 77.5| 73.8 kh/s
[2013-05-01 21:35:14]  10: 15.4| 30.8| 46.2| 49.2| 76.9| 61.5| 86.2| 82.1 kh/s
[2013-05-01 21:35:14]  11: 16.9| 33.8| 50.8| 54.2| 84.6| 81.2| 79.0| 90.3 kh/s
[2013-05-01 21:35:14] LONGPOLL detected new block
[2013-05-01 21:35:14] DEBUG: got new work
[2013-05-01 21:35:15]  12: 18.5| 36.9| 55.4| 59.1| 92.3| 73.8|103.4| 98.5 kh/s
[2013-05-01 21:35:15]  13: 20.0| 40.0| 48.0| 80.0| 80.0| 96.0|109.2|106.7 kh/s
[2013-05-01 21:35:16]  14: 21.5| 43.1| 51.7| 68.9| 86.2|103.4|120.6|114.9 kh/s
[2013-05-01 21:35:16]  15: 23.1| 46.2| 55.4| 73.8| 76.9                   kh/s
[2013-05-01 21:35:17]  16: 24.6| 49.2| 59.1| 78.8| 82.1                   kh/s
[2013-05-01 21:35:17]  17: 26.2| 52.3| 62.8| 83.7| 87.2                   kh/s
[2013-05-01 21:35:17]  18: 27.7| 55.4| 66.5| 88.6| 92.3                   kh/s
[2013-05-01 21:35:18]  19: 29.2| 58.5| 87.7| 77.9| 97.4                   kh/s
[2013-05-01 21:35:18]  20: 30.8| 61.5| 73.8| 98.5|102.6                   kh/s
[2013-05-01 21:35:19]  21: 32.3| 64.6| 64.6|103.4|107.7                   kh/s
[2013-05-01 21:35:19]  22: 33.8| 67.7| 67.7| 90.3                         kh/s
[2013-05-01 21:35:19]  23: 35.4| 70.8| 70.8| 94.4                         kh/s
[2013-05-01 21:35:20]  24: 36.9| 73.8| 88.6| 98.5                         kh/s
[2013-05-01 21:35:20]  25: 38.5| 76.9| 92.3|102.6                         kh/s
[2013-05-01 21:35:21]  26: 40.0| 80.0| 80.0|106.7                         kh/s
[2013-05-01 21:35:21]  27: 41.5| 83.1| 83.1|110.8                         kh/s
[2013-05-01 21:35:21]  28: 43.1| 86.2| 86.2|114.9                         kh/s
[2013-05-01 21:35:22]  29: 44.6| 89.2| 89.2                               kh/s
[2013-05-01 21:35:22]  30: 46.2| 92.3| 92.3                               kh/s
[2013-05-01 21:35:22]  31: 47.7| 95.4| 95.4                               kh/s
[2013-05-01 21:35:22]  32: 49.2| 98.5| 98.5                               kh/s
[2013-05-01 21:35:23]  33: 50.8|101.5|101.5                               kh/s
[2013-05-01 21:35:23]  34: 52.3| 83.7|104.6                               kh/s
[2013-05-01 21:35:23]  35: 53.8| 71.8|107.7                               kh/s
[2013-05-01 21:35:23]  36: 55.4| 73.8                                     kh/s
[2013-05-01 21:35:24]  37: 56.9| 75.9                                     kh/s
[2013-05-01 21:35:24]  38: 58.5| 77.9                                     kh/s
[2013-05-01 21:35:24]  39: 48.0| 96.0                                     kh/s
[2013-05-01 21:35:24]  40: 49.2| 98.5                                     kh/s
[2013-05-01 21:35:25]  41: 63.1|100.9                                     kh/s
[2013-05-01 21:35:25]  42: 64.6| 86.2                                     kh/s
[2013-05-01 21:35:25]  43: 66.2| 88.2                                     kh/s
[2013-05-01 21:35:25]  44: 54.2|108.3                                     kh/s
[2013-05-01 21:35:26]  45: 69.2| 92.3                                     kh/s
[2013-05-01 21:35:26]  46: 70.8| 94.4                                     kh/s
[2013-05-01 21:35:26]  47: 57.8|115.7                                     kh/s
[2013-05-01 21:35:26]  48: 73.8| 98.5                                     kh/s
[2013-05-01 21:35:26]  49: 60.3|120.6                                     kh/s
[2013-05-01 21:35:27]  50: 76.9|102.6                                     kh/s
[2013-05-01 21:35:27]  51: 78.5|104.6                                     kh/s
[2013-05-01 21:35:27]  52: 80.0|106.7                                     kh/s
[2013-05-01 21:35:27]  53: 81.5|108.7                                     kh/s
[2013-05-01 21:35:27]  54: 83.1|110.8                                     kh/s
[2013-05-01 21:35:28]  55: 84.6|112.8                                     kh/s
[2013-05-01 21:35:28]  56: 86.2|114.9                                     kh/s
[2013-05-01 21:35:28]    S   x1    x2    x3    x4    x5    x6    x7    x8
[2013-05-01 21:35:28]   1:  1.5|  3.1|  4.6|  4.9|  7.7|  7.4| 10.8|  8.2 kh/s
[2013-05-01 21:35:29]   2:  3.1|  6.2|  9.2|  9.8| 15.4| 14.8| 17.2| 19.7 kh/s
[2013-05-01 21:35:30]   3:  4.6|  9.2| 13.8| 14.8| 23.1| 18.5| 32.3| 24.6 kh/s
[2013-05-01 21:35:30]   4:  6.2| 12.3| 18.5| 19.7| 30.8| 24.6| 34.5| 39.4 kh/s
[2013-05-01 21:35:31]   5:  7.7| 15.4| 18.5| 30.8| 30.8| 46.2| 35.9| 49.2 kh/s
[2013-05-01 21:35:31]   6:  9.2| 18.5| 22.2| 29.5| 46.2| 36.9| 51.7| 59.1 kh/s
[2013-05-01 21:35:32]   7: 10.8| 21.5| 25.8| 34.5| 43.1| 51.7| 60.3| 68.9 kh/s
[2013-05-01 21:35:32]   8: 12.3| 24.6| 29.5| 39.4| 49.2| 49.2| 57.4| 65.6 kh/s
[2013-05-01 21:35:33]   9: 13.8| 22.2| 33.2| 44.3| 55.4| 55.4| 64.6| 73.8 kh/s
[2013-05-01 21:35:33]  10: 15.4| 24.6| 36.9| 49.2| 61.5| 61.5| 71.8| 82.1 kh/s
[2013-05-01 21:35:34]  11: 16.9| 22.6| 50.8| 45.1| 67.7| 67.7| 79.0| 90.3 kh/s
[2013-05-01 21:35:35]  12: 18.5| 29.5| 44.3| 59.1| 73.8| 73.8| 86.2| 98.5 kh/s
[2013-05-01 21:35:35]  13: 20.0| 40.0| 40.0| 64.0| 66.7| 96.0| 93.3| 85.3 kh/s
[2013-05-01 21:35:36]  14: 21.5| 34.5| 51.7| 57.4| 86.2| 86.2|100.5|114.9 kh/s
[2013-05-01 21:35:36]  15: 18.5| 46.2| 46.2| 73.8                         kh/s
[2013-05-01 21:35:36]  16: 24.6| 49.2| 49.2| 65.6| 82.1                   kh/s
[2013-05-01 21:35:37]  17: 26.2| 41.8| 62.8| 69.7| 87.2                   kh/s
[2013-05-01 21:35:37]  18: 27.7| 44.3| 55.4| 88.6| 92.3                   kh/s
[2013-05-01 21:35:38]  19: 29.2| 46.8| 58.5| 93.5| 97.4                   kh/s
[2013-05-01 21:35:38]  20: 30.8| 41.0| 73.8| 82.1|102.6                   kh/s
[2013-05-01 21:35:39]  21: 32.3| 51.7| 64.6| 86.2|107.7                   kh/s
[2013-05-01 21:35:39]  22: 33.8| 45.1| 81.2| 90.3                         kh/s
[2013-05-01 21:35:39]  23: 35.4| 56.6| 70.8| 94.4                         kh/s
[2013-05-01 21:35:40]  24: 36.9| 59.1| 73.8| 98.5                         kh/s
[2013-05-01 21:35:40]  25: 38.5| 61.5| 76.9|102.6                         kh/s
[2013-05-01 21:35:40]  26: 40.0| 64.0| 80.0|106.7                         kh/s
[2013-05-01 21:35:40]  27: 41.5| 55.4| 99.7                               kh/s
[2013-05-01 21:35:41]  28: 43.1| 68.9| 86.2|114.9                         kh/s
[2013-05-01 21:35:41]  29: 44.6| 59.5| 89.2                               kh/s
[2013-05-01 21:35:41]  30: 46.2| 61.5| 92.3                               kh/s
[2013-05-01 21:35:42]  31: 47.7| 76.3| 95.4                               kh/s
[2013-05-01 21:35:42]  32: 49.2| 65.6| 98.5                               kh/s
[2013-05-01 21:35:42]  33: 50.8| 67.7|101.5                               kh/s
[2013-05-01 21:35:42]  34: 52.3| 69.7|104.6                               kh/s
[2013-05-01 21:35:43]  35: 53.8| 71.8|107.7                               kh/s
[2013-05-01 21:35:43]  36: 44.3| 88.6                                     kh/s
[2013-05-01 21:35:43]  37: 56.9| 75.9                                     kh/s
[2013-05-01 21:35:43]  38: 58.5| 77.9                                     kh/s
[2013-05-01 21:35:43]  39: 48.0| 96.0                                     kh/s
[2013-05-01 21:35:44]  40: 49.2| 98.5                                     kh/s
[2013-05-01 21:35:44]  41: 63.1| 84.1                                     kh/s
[2013-05-01 21:35:44]  42: 64.6| 86.2                                     kh/s
[2013-05-01 21:35:44]  43: 52.9| 88.2                                     kh/s
[2013-05-01 21:35:44]  44: 67.7| 90.3                                     kh/s
[2013-05-01 21:35:45]  45: 55.4| 92.3                                     kh/s
[2013-05-01 21:35:45]  46: 56.6| 94.4                                     kh/s
[2013-05-01 21:35:45]  47: 72.3| 96.4                                     kh/s
[2013-05-01 21:35:45]  48: 59.1| 98.5                                     kh/s
[2013-05-01 21:35:45]  49: 75.4|100.5                                     kh/s
[2013-05-01 21:35:46]  50: 61.5|102.6                                     kh/s
[2013-05-01 21:35:46]  51: 78.5                                           kh/s
[2013-05-01 21:35:46]  52: 80.0                                           kh/s
[2013-05-01 21:35:46]  53: 81.5                                           kh/s
[2013-05-01 21:35:46]  54: 66.5|110.8                                     kh/s
[2013-05-01 21:35:46]  55: 84.6                                           kh/s
[2013-05-01 21:35:47]  56: 86.2                                           kh/s
[2013-05-01 21:35:47] GPU #0:  120.62 khash/s with configuration  14x7
[2013-05-01 21:35:47] GPU #0: using launch configuration  14x7
KnowBuddy
Member
**
Offline Offline

Activity: 69
Merit: 10


View Profile
May 02, 2013, 05:23:21 AM
Last edit: May 02, 2013, 07:06:47 AM by KnowBuddy
 #680

This is written in the Readme, hope it helps:

Currently there is just one prefix, which is "S". Later releases may
see the introduction of more kernel variants with using other letters.

Examples:

e.g. S27x3 is a launch configuration that works well on GTX 260
      28x4 is a launch configuration that works on Geforce GTX 460
     290x2 is a launch configuration that works on Geforce GTX 660Ti

You should wait through autotune to see what kernel is found best for
your current hardware configuration.

The choice between Non-Titan and Titan CUDA kernels is automatically
made based on your device's compute capability. Titans cost around
a thousand dollars, so you probably don't have one.


Prefix  | Non-Titan          | Titan
-------------------------------------------------------
 <none> | low shared memory  | default kernel
        | optimized kernel   | with funnel shifter
        |                    |
   S    | spinlock kernel    | spinlock kernel
        | for Kepler GPUs    | with funnel shifter

Can anyone explain like i'm 5? If my card used to autotune to the default kernal and is now auto tuning to the spinlock kernel will it actually make a difference?

Here's a quote from Christian which goes into more detail about the -l flag format that helped me understand it much better than the README:

Can someone explain this "64x2" "S27x3,28x4" thing to me or point me in the right direction on reading up on it??

I have a 580gtx and i'm trying to figure out the best set up

well i did not figure out the meaning, but if you run it and let it autotune, it will choose automatically what's the best one (then you can add the flag in the batch file, like -l 112x2 for me)

112x2 means it throws 112 blocks at CUDA, and each consists of 2 warps. A warp is a group of 32 threads.

So in total it computes 112*2*32 = 7168 hashes in parallel in a single CUDA kernel launch.

And because the scrypt scratchpad is 131072 bytes long, this would consume 7168*131072 bytes of memory
on the card. That's about 917 MB.

So, as I understand it...

The kernel selection options (by autotune or -l flag) prior to the 2013-04-30 release were of the format:
  • "S" OR ""(no value)] - where S is optimized for older devices (compute capability < 2.0?) and no value is for all other devices
  • #b - # of blocks
  • x
  • #w - # of warps (groups of 32 threads)

And the kernel selection options (by autotune or -l flag) for the 2013-04-30 release onward are of the format:
  • "S" OR ""(no value)] - where S is optimized for Kepler devices (some GTX 6xx GPUs) and no value is for all other devices
  • #b - # of blocks
  • x
  • #w - # of warps (groups of 32 threads)

An important consideration is that #b x #w x 32 x 131072 should be less than the RAM (in bytes) on the card in question.

I'm not really sure how to break it down to explain-like-I'm-five levels without resorting to some MS paint-esque diagramming, which probably would do more harm than good.

And the answer to your question is... maybe. I have a Kepler mobile card but I saw equivalent performance when switching on the S flag at the same blocks and warps, and texture caching settings. I haven't had a chance to run any extensive testing or autotuning yet though.
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 [34] 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 ... 1135 »
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!