Bitcoin Forum
May 03, 2024, 05:51:38 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: « 1 ... 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 [132] 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 ... 233 »
  Print  
Author Topic: [ANN] sgminer v5 - optimized X11/X13/NeoScrypt/Lyra2RE/etc. kernel-switch miner  (Read 877795 times)
platinum4
Sr. Member
****
Offline Offline

Activity: 547
Merit: 250



View Profile
December 26, 2014, 01:12:09 AM
 #2621

What are you guys getting for Neoscrypt efficiency? Hash per watt, power draw at the wall?
Dick.  But it appears to be most profitable on nicehash again; even with the salty bottlenecked neoscrypt.cl you released for us around Veteran's Day.  (thanks again)

Actually, that was the best I had at the time - also, you aren't entitled to jack shit from me, need I remind you? Complain to Vehre - his code got less than 100kh/s on a 280X. Would you rather have that instead?

For the rest of you, since I've gotten a lot of people asking, I made another SGMiner 5 build off the git master branch: https://litecointalk.org/index.php?topic=17533 -- see the bottom of the OP for the latest.

Did you ever get around to identifying your neoscrypt kernels so that people could tell which is which?

Well, now there's really just one major one, and the Hawaii one that wasn't a good idea.

Well, it wasn't until I downloaded the new SG5-wolf from yesterday and compared the neoscrypt files that I found out that it had the same 2 month old kernel..  For some reason I was expecting a different neoscrypt kernel in it.

I think that's just a 64-bit Windows build as opposed to 32-bit ones from badman74, djm34, etc.
1714758698
Hero Member
*
Offline Offline

Posts: 1714758698

View Profile Personal Message (Offline)

Ignore
1714758698
Reply with quote  #2

1714758698
Report to moderator
1714758698
Hero Member
*
Offline Offline

Posts: 1714758698

View Profile Personal Message (Offline)

Ignore
1714758698
Reply with quote  #2

1714758698
Report to moderator
1714758698
Hero Member
*
Offline Offline

Posts: 1714758698

View Profile Personal Message (Offline)

Ignore
1714758698
Reply with quote  #2

1714758698
Report to moderator
Whoever mines the block which ends up containing your transaction will get its fee.
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
1714758698
Hero Member
*
Offline Offline

Posts: 1714758698

View Profile Personal Message (Offline)

Ignore
1714758698
Reply with quote  #2

1714758698
Report to moderator
1714758698
Hero Member
*
Offline Offline

Posts: 1714758698

View Profile Personal Message (Offline)

Ignore
1714758698
Reply with quote  #2

1714758698
Report to moderator
1714758698
Hero Member
*
Offline Offline

Posts: 1714758698

View Profile Personal Message (Offline)

Ignore
1714758698
Reply with quote  #2

1714758698
Report to moderator
damm315er
Sr. Member
****
Offline Offline

Activity: 539
Merit: 255


View Profile
December 26, 2014, 02:48:59 AM
 #2622

What are you guys getting for Neoscrypt efficiency? Hash per watt, power draw at the wall?
Dick.  But it appears to be most profitable on nicehash again; even with the salty bottlenecked neoscrypt.cl you released for us around Veteran's Day.  (thanks again)

Actually, that was the best I had at the time - also, you aren't entitled to jack shit from me, need I remind you? Complain to Vehre - his code got less than 100kh/s on a 280X. Would you rather have that instead?

For the rest of you, since I've gotten a lot of people asking, I made another SGMiner 5 build off the git master branch: https://litecointalk.org/index.php?topic=17533 -- see the bottom of the OP for the latest.

Did you ever get around to identifying your neoscrypt kernels so that people could tell which is which?

Well, now there's really just one major one, and the Hawaii one that wasn't a good idea.

Well, it wasn't until I downloaded the new SG5-wolf from yesterday and compared the neoscrypt files that I found out that it had the same 2 month old kernel..  For some reason I was expecting a different neoscrypt kernel in it.

I think that's just a 64-bit Windows build as opposed to 32-bit ones from badman74, djm34, etc.

Yeah - plus no DLLs and such to carry around.

Yeah, those are heavy...   lol
badman74
Hero Member
*****
Offline Offline

Activity: 658
Merit: 500



View Profile
December 26, 2014, 03:58:39 AM
 #2623

What are you guys getting for Neoscrypt efficiency? Hash per watt, power draw at the wall?
Dick.  But it appears to be most profitable on nicehash again; even with the salty bottlenecked neoscrypt.cl you released for us around Veteran's Day.  (thanks again)

Actually, that was the best I had at the time - also, you aren't entitled to jack shit from me, need I remind you? Complain to Vehre - his code got less than 100kh/s on a 280X. Would you rather have that instead?

For the rest of you, since I've gotten a lot of people asking, I made another SGMiner 5 build off the git master branch: https://litecointalk.org/index.php?topic=17533 -- see the bottom of the OP for the latest.

Did you ever get around to identifying your neoscrypt kernels so that people could tell which is which?

Well, now there's really just one major one, and the Hawaii one that wasn't a good idea.

Well, it wasn't until I downloaded the new SG5-wolf from yesterday and compared the neoscrypt files that I found out that it had the same 2 month old kernel..  For some reason I was expecting a different neoscrypt kernel in it.

I think that's just a 64-bit Windows build as opposed to 32-bit ones from badman74, djm34, etc.

Yeah - plus no DLLs and such to carry around.
I would love to build it to work without the dll's but I have no idea how...
damm315er
Sr. Member
****
Offline Offline

Activity: 539
Merit: 255


View Profile
December 26, 2014, 04:51:41 AM
 #2624

What are you guys getting for Neoscrypt efficiency? Hash per watt, power draw at the wall?
Dick.  But it appears to be most profitable on nicehash again; even with the salty bottlenecked neoscrypt.cl you released for us around Veteran's Day.  (thanks again)

Actually, that was the best I had at the time - also, you aren't entitled to jack shit from me, need I remind you? Complain to Vehre - his code got less than 100kh/s on a 280X. Would you rather have that instead?

For the rest of you, since I've gotten a lot of people asking, I made another SGMiner 5 build off the git master branch: https://litecointalk.org/index.php?topic=17533 -- see the bottom of the OP for the latest.

Did you ever get around to identifying your neoscrypt kernels so that people could tell which is which?

Well, now there's really just one major one, and the Hawaii one that wasn't a good idea.

Well, it wasn't until I downloaded the new SG5-wolf from yesterday and compared the neoscrypt files that I found out that it had the same 2 month old kernel..  For some reason I was expecting a different neoscrypt kernel in it.

I think that's just a 64-bit Windows build as opposed to 32-bit ones from badman74, djm34, etc.

Yeah - plus no DLLs and such to carry around.
I would love to build it to work without the dll's but I have no idea how...

Static linking is awesome. I also GPG sign my builds, so I'm staking my reputation on both the forums and OTC that anything I sign is clean.

I've seen those pics of your desktop, what kind of a reputation are you trying to establish?  lol
kenshirothefist (OP)
Sr. Member
****
Offline Offline

Activity: 457
Merit: 273



View Profile
December 26, 2014, 08:51:00 AM
 #2625

@Wolf0, @badman74, @djm34, etc. guys, guys, guys ... as I said many times before you are all great programmers (and I sent tips to all of you for your great public work) but you really lack one skill that is crucial for open source development, and that is team work. It is OK for each of you to optimize some kernels and not share it publicly to get maximum profit out of your own work. But as far as core development and binary distribution it should be done at one central place. People on this thread get lost when looking for binaries, trying various builds, many times hits an older build, build with bugs, etc. Linus Torvalds designed and developed Git with powerful branching in mind with this purpose - you can keep your branches, but should also maintain a master branch and not make thousands of forks even for smallest and marginal changes. https://github.com/sgminer-dev/sgminer is a good central place and binaries could be placed at the Releases section https://github.com/sgminer-dev/sgminer/releases ... So, can you put your self together and work as a team?

You should all take a look at this talk: https://www.youtube.com/watch?v=5pOxlazS3zs (especially from minute 2 to minute 5, but preferably whole video).

Anyway, keep up the good work Wink
djm34
Legendary
*
Offline Offline

Activity: 1400
Merit: 1050


View Profile WWW
December 26, 2014, 11:47:00 AM
 #2626

@Wolf0, @badman74, @djm34, etc. guys, guys, guys ... as I said many times before you are all great programmers (and I sent tips to all of you for your great public work) but you really lack one skill that is crucial for open source development, and that is team work. It is OK for each of you to optimize some kernels and not share it publicly to get maximum profit out of your own work. But as far as core development and binary distribution it should be done at one central place. People on this thread get lost when looking for binaries, trying various builds, many times hits an older build, build with bugs, etc. Linus Torvalds designed and developed Git with powerful branching in mind with this purpose - you can keep your branches, but should also maintain a master branch and not make thousands of forks even for smallest and marginal changes. https://github.com/sgminer-dev/sgminer is a good central place and binaries could be placed at the Releases section https://github.com/sgminer-dev/sgminer/releases ... So, can you put your self together and work as a team?

You should all take a look at this talk: https://www.youtube.com/watch?v=5pOxlazS3zs (especially from minute 2 to minute 5, but preferably whole video).

Anyway, keep up the good work Wink
nice presentation... however I don't remember what was exactly between 2 and 5min. (also my interest in linux stopped a long time ago when they tried to make it looks like windows with ugly icons and get away from the unix looks and feel...  and even made things difficult just to open a terminal... at that point I preferred to switch entirely to windows)

I intended to move my change to sgminer-dev (even made a pull request) but I was asked to make a few change to comply with their coding rule.... and I don't have time for this at the moment, but I will come to it eventually...

djm34 facebook page
BTC: 1NENYmxwZGHsKFmyjTc5WferTn5VTFb7Ze
Pledge for neoscrypt ccminer to that address: 16UoC4DmTz2pvhFvcfTQrzkPTrXkWijzXw
bitspender
Hero Member
*****
Offline Offline

Activity: 658
Merit: 500



View Profile
December 26, 2014, 05:13:40 PM
 #2627

What's the recommended settings for a R9 290 trixx ?
Only getting 110Kh on neoscrypt  Angry
andro1970
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
December 26, 2014, 05:47:58 PM
 #2628

I have some problem ....

[23:30:51] Started sgminer 4.2.2-310-g8985f-dirty
[23:30:51] * using Jansson 2.6
[23:30:51] Loaded configuration file sgminer.conf
[23:30:51] Probing for an alive pool
[23:30:52] Switching to NiceHash_X13_multi - first alive pool
[23:30:52] Startup GPU initialization... Using settings from pool NiceHash_X13_multi.
[23:30:52] Startup Pool No = 3
[23:30:52] Initialising kernel marucoin-mod.cl with bitalign, unpatched BFI, nfactor 10, n 1024
[23:30:52] Maximum buffer memory device 0 supports says 535822336
[23:30:52] Your settings come to 536870912
[23:30:52] Error -61: clCreateBuffer (padbuffer8), decrease TC or increase LG
[23:30:52] Failed to init GPU thread 0, disabling device 0
[23:30:52] Restarting the GPU from the menu will not fix this.
[23:30:52] Re-check your configuration and try restarting.
Press enter to continue:

p.s
set GPU_MAX_ALLOC_PERCENT=100 - enable
CPU:AMD-A4-6300 with Radeon
RAM:4G
damm315er
Sr. Member
****
Offline Offline

Activity: 539
Merit: 255


View Profile
December 26, 2014, 05:53:41 PM
 #2629

What's the recommended settings for a R9 290 trixx ?
Only getting 110Kh on neoscrypt  Angry

Run SGminer with the latest kernel, with 14.9 drivers on the machine and 14.6 drivers in the folder.  And pick your favorite intensity/xintensity settings.  

With the latest bottlenecked kernel, I've gotten 330k+ each, detuned a little because of bad fans. (the 330k+ are trixx 290's and the HW's are because of heat, the 301k are other GPU's)

andro1970
Newbie
*
Offline Offline

Activity: 4
Merit: 0


View Profile
December 26, 2014, 05:57:59 PM
 #2630

I have some problem ....

[23:30:51] Started sgminer 4.2.2-310-g8985f-dirty
[23:30:51] * using Jansson 2.6
[23:30:51] Loaded configuration file sgminer.conf
[23:30:51] Probing for an alive pool
[23:30:52] Switching to NiceHash_X13_multi - first alive pool
[23:30:52] Startup GPU initialization... Using settings from pool NiceHash_X13_multi.
[23:30:52] Startup Pool No = 3
[23:30:52] Initialising kernel marucoin-mod.cl with bitalign, unpatched BFI, nfactor 10, n 1024
[23:30:52] Maximum buffer memory device 0 supports says 535822336
[23:30:52] Your settings come to 536870912
[23:30:52] Error -61: clCreateBuffer (padbuffer8), decrease TC or increase LG
[23:30:52] Failed to init GPU thread 0, disabling device 0
[23:30:52] Restarting the GPU from the menu will not fix this.
[23:30:52] Re-check your configuration and try restarting.
Press enter to continue:

p.s
set GPU_MAX_ALLOC_PERCENT=100 - enable
CPU:AMD-A4-6300 with Radeon
RAM:4G

And you don't bother to show us the command line/batch file/conf file...

Code:
{
"pools" : [
     {
          "name" : "NiceHash_Scrypt_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4333",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "10",
          "algorithm" : "zuikkis"
     },
     {
          "name" : "NiceHash_Scrypt-N_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4335",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "11",
          "algorithm" : "zuikkis"
     },
     {
          "name" : "NiceHash_X11_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4336",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "10",
          "algorithm" : "darkcoin-mod"
     },
     {
          "name" : "NiceHash_X13_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4337",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "10",
          "algorithm" : "marucoin-mod"
     },
     {
          "name" : "NiceHash_Keccak_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4338",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "10",
          "algorithm" : "maxcoin"
     },
     {
          "name" : "NiceHash_X15_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4339",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "10",
          "algorithm" : "bitblock"
     },
     {
          "name" : "NiceHash_NIST5_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4340",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "10",
          "algorithm" : "talkcoin-mod"
     },
     {
          "name" : "NiceHash_NeoScrypt_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4341",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "10",
          "algorithm" : "neoscrypt"
     },
     {
          "name" : "NiceHash_Lyra2RE_multi",
          "url" : "stratum+tcp://stratum.nicehash.com:4342",
          "user" : "1hajStUB86VJwBoV7C5yodGTaW3chxstA",
          "pass" : "x",
          "nfactor" : "10",
          "algorithm" : "Lyra2RE"
     }

  ],
"failover-only" : true,
"failover-switch-delay" : "300",
"queue" : "0"
}

command line is empty (only sdminer.exe)
bitspender
Hero Member
*****
Offline Offline

Activity: 658
Merit: 500



View Profile
December 26, 2014, 09:33:38 PM
 #2631

What's the recommended settings for a R9 290 trixx ?
Only getting 110Kh on neoscrypt  Angry

Run SGminer with the latest kernel, with 14.9 drivers on the machine and 14.6 drivers in the folder.  And pick your favorite intensity/xintensity settings.  

With the latest bottlenecked kernel, I've gotten 330k+ each, detuned a little because of bad fans. (the 330k+ are trixx 290's and the HW's are because of heat, the 301k are other GPU's)



What kernel is it?
damm315er
Sr. Member
****
Offline Offline

Activity: 539
Merit: 255


View Profile
December 26, 2014, 10:27:33 PM
 #2632

What's the recommended settings for a R9 290 trixx ?
Only getting 110Kh on neoscrypt  Angry

Run SGminer with the latest kernel, with 14.9 drivers on the machine and 14.6 drivers in the folder.  And pick your favorite intensity/xintensity settings.  

With the latest bottlenecked kernel, I've gotten 330k+ each, detuned a little because of bad fans. (the 330k+ are trixx 290's and the HW's are because of heat, the 301k are other GPU's)



What kernel is it?

What makes you think it's not the one that comes with SGMiner? (It is, I think.)

It is the kernel that comes with the current SGminer...  Either the nicehash dev build, or the wolf branch, it makes no difference.  (although the 64 bit wolf version is good for about an extra ~2 kh/s per GPU, being 64 bit)

I did specify the latest kernel.
bitspender
Hero Member
*****
Offline Offline

Activity: 658
Merit: 500



View Profile
December 27, 2014, 04:22:00 AM
 #2633

What's the recommended settings for a R9 290 trixx ?
Only getting 110Kh on neoscrypt  Angry

Try xintensity 3 or 4 with 2 gpu threads, also, overclock it.

Latest SGminer, 14.9 installed using 14.6 drivers.

xintensity gives me HW errors (nonce)
intensity 14-17 gives me only 237Kh/s

Any tips for 290 trixx?

Code:
"pools" : [

{
"name" : "",
"url" : "",
"user" : "",
"pass" : "",
"intensity" : "17",
"kernel" : "neoscrypt"
}
],
"failover-only" : true,
"failover-switch-delay" : "300",
"queue" : "0",

"api-listen": true,
"api-port": "4028",
"api-allow": "W:127.0.0.1",

"auto-fan" : true,
"auto-gpu" : true,
"gpu-threads" : "2",
"kernel-path" : "/usr/local/bin"
}
edonkey
Legendary
*
Offline Offline

Activity: 1150
Merit: 1004



View Profile
December 27, 2014, 04:42:05 AM
 #2634

do you plan to make a build for mac os?

I'm interested in a Mac OS X version too.

Was I helpful?   BTC: 3G1Ubof5u8K9iJkM8We2f3amYZgGVdvpHr
bitspender
Hero Member
*****
Offline Offline

Activity: 658
Merit: 500



View Profile
December 27, 2014, 04:52:59 AM
 #2635

Try 13, like damm315er is using.

Thx wolf, 13 is giving me 240. Still far away from +300
bitspender
Hero Member
*****
Offline Offline

Activity: 658
Merit: 500



View Profile
December 27, 2014, 04:59:51 AM
 #2636

Try 13, like damm315er is using.

Thx wolf, 13 is giving me 240. Still far away from +300

Try xI 3 or 4, but specify thread concurrency with it. Set it to 16384.

Also, what clocks?

xi 3-4 with concurrency 16384 gives me even less, around 220.
Clocks default 1000-1300, tried on 1050-1400 to.
damm315er
Sr. Member
****
Offline Offline

Activity: 539
Merit: 255


View Profile
December 27, 2014, 09:56:36 AM
 #2637

Try 13, like damm315er is using.

Thx wolf, 13 is giving me 240. Still far away from +300

Try xI 3 or 4, but specify thread concurrency with it. Set it to 16384.

Also, what clocks?

xi 3-4 with concurrency 16384 gives me even less, around 220.
Clocks default 1000-1300, tried on 1050-1400 to.

try 1130/1475, with -I 13 & -g 2,  and keep tinkering with the worksize and TC.  It took me about 20 hours of trial and error to hit that mark.

Speaking of which, those GPU's just came back from warranty replacement, time to see if I can top it with good fans!!
Zuikkis
Newbie
*
Offline Offline

Activity: 57
Merit: 0


View Profile
December 27, 2014, 10:54:09 AM
 #2638

Try memclock 1500, or if that is not stable, try 1250.

AMD gpus  have memory  timings set at 125MHz intervals, always ending at even 125MHz. Like this:

875-1000
1001-1125
1126-1250
1251-1375
1376-1500
1501-1625

 mining is mostly about random access latency, not about sequential reads. You get best random access speed at the high end of each range, like 1250 or 1500.

It's very likely that 1250MHz is faster than 1400MHz, because 1400MHz is so close to the start of the range and 1250 is exactly at the end..
bitspender
Hero Member
*****
Offline Offline

Activity: 658
Merit: 500



View Profile
December 27, 2014, 05:44:11 PM
 #2639

Got it running around 310Kh.

I think i'll take that instead of tweaking further to gain just a little bit more.

Thank you all!
jch9678
Full Member
***
Offline Offline

Activity: 169
Merit: 100


View Profile
December 27, 2014, 07:35:58 PM
 #2640

Try 13, like damm315er is using.

Thx wolf, 13 is giving me 240. Still far away from +300

Try xI 3 or 4, but specify thread concurrency with it. Set it to 16384.

Also, what clocks?

I know this was a while ago but ystarnaud said that xintensity and rawintensity are unused. 

https://bitcointalk.org/index.php?topic=632503.msg9475388#msg9475388

That was on Nov. 7 so I guess xI does now work with neo.  That's good news I was about to make a bunch of bins with different intensities with the 13.12 drivers and then upgrade to 14.6 rc2.  Is making bins with 13.12 a waste of time?

BTC: 15GqpmqNNJ1REWrDWTfymh7moos1sEvz7A
Pages: « 1 ... 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 [132] 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 ... 233 »
  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!