[ANN] sgminer v5 - optimized X11/X13/NeoScrypt/Lyra2RE/etc. kernel-switch miner - page 236.

cgladue

member

Activity: 97

Merit: 10

Quote from: kenshirothefist on June 11, 2014, 06:03:42 AM

Today's fresh builds with many bugfixes available for download: https://www.nicehash.com/software/#sgminer

when i downloaded the win32 build and extracted/ran it there was no "kernel" folder and sgminer refused to start. i had to move all the .cl files to a folder named kernel manually before it would run, just a heads up

xsoft

newbie

Activity: 60

Merit: 0

Quote from: kenshirothefist on June 11, 2014, 06:03:42 AM

Today's fresh builds with many bugfixes available for download: https://www.nicehash.com/software/#sgminer

Will do thanks.
Btw, Ive remove TC param from root config and let just G=1 at pool.scrypt.n part which helped.
It seems that unique TC per pool was ignored if global G was set. (tested on sgminer-5.0-pre-release-2014-06-06_01-linux-i386.zip)

kenshirothefist

sr. member

Activity: 457

Merit: 273

Today's fresh builds with many bugfixes available for download: https://www.nicehash.com/software/#sgminer

kenshirothefist

sr. member

Activity: 457

Merit: 273

Quote from: xsoft on June 11, 2014, 05:54:04 AM

[10:48:06] Error -4: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel)

This is usually out-of-memory issue. Try lowering TC to 4096.

xsoft

newbie

Activity: 60

Merit: 0

Hello,
I still cant make Scrypt-N working on 7870 (R270x) @moded BAMT 1.2, sgminer 5.0 (from NiceHash)
Other algor. are working (scrypt, x11, x13, keccak).

Im trying TC 8192 or 125168
engine 1020, memory 1250
and also few other modifications.

Ive tried also xIntensity 1.
I thing that gpu-threads are not making any evect. Im seeng "g2" at .bin build even if set to g=1.

Code:

sgminer 4.2.1 - Started: [2014-06-11 10:47:47] - [0 days 00:00:26]
--------------------------------------------------------------------------------
(5s):0.000 (avg):0.000h/s | A:0  R:0  HW:0  WU:106.714/m
ST: 2  SS: 0  NB: 4  LW: 20  GF: 0  RF: 0
Connected to NiceHash_Scrypt-N (stratum) diff 256 as user 1xxxxxxxxxxxxxxxxxxxxxxxxxxxx
Block: fa157e63...  Diff:17.4K  Started: [10:47:54]  Best share: 7.038
--------------------------------------------------------------------------------
[P]ool management [G]PU management [S ]ettings [D]isplay options [Q]uit
GPU 0:  46.0C 2139RPM | OFF   /58.14Kh/s | R:  0.0% HW:0 WU:57.461/m xI:  4
GPU 1:  44.0C 2123RPM | OFF   /58.49Kh/s | R:  0.0% HW:0 WU:53.357/m xI:  4
--------------------------------------------------------------------------------
[10:47:56] Initialising kernel zuikkis.cl with bitalign, unpatched BFI, nfactor 10, n 1024
[10:47:56] Initialising kernel zuikkis.cl with bitalign, unpatched BFI, nfactor 10, n 1024
[10:47:56] Initialising kernel zuikkis.cl with bitalign, unpatched BFI, nfactor 10, n 1024
[10:47:56] Initialising kernel zuikkis.cl with bitalign, unpatched BFI, nfactor 10, n 1024
[10:47:56] Building binary zuikkisPitcairnglg2tc8000nf11w256l4.bin
[10:48:06] Initialising kernel zuikkis.cl with bitalign, unpatched BFI, nfactor 11, n 2048
[10:48:06] Initialising kernel zuikkis.cl with bitalign, unpatched BFI, nfactor 11, n 2048
[10:48:06] Error -4: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel)
[10:48:06] GPU 0 failure, disabling!
[10:48:06] Initialising kernel zuikkis.cl with bitalign, unpatched BFI, nfactor 11, n 2048
[10:48:06] Initialising kernel zuikkis.cl with bitalign, unpatched BFI, nfactor 11, n 2048
[10:48:06] Error -4: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel)
[10:48:06] GPU 1 failure, disabling!
[10:48:11] Thread 0 being disabled
[10:48:11] Thread 3 being disabled

Code:

{
"pools" : [
        {
                "name" : "NiceHash_Scrypt",
                "url" : "stratum+tcp://stratum.nicehash.com:3333",
                "user" : "1xxxxxxxxxxxxxxxxxxxxxxxxxxxx",
                "pass" : "x",
                "pool-algorithm" : "zuikkis",
                "pool-nfactor" : "10",
                "pool-xintensity" : "4",
                "pool-thread-concurrency" : "8192",
                "pool-gpu-threads" : "2"
        },
        {
                "name" : "NiceHash_Scrypt-N", <===========================
                "url" : "stratum+tcp://stratum.nicehash.com:3335",
                "user" : "1xxxxxxxxxxxxxxxxxxxxxxxxxxxx",
                "pass" : "x",
                "pool-algorithm" : "zuikkis",
                "pool-nfactor" : "11",
                "pool-xintensity" : "4",
                "pool-thread-concurrency" : "8000",
                "pool-gpu-threads" : "2",
                "pool-gpu-engine" : "1020",
                "pool-gpu-memclock" : "1250"
        },
        {
                "name" : "NiceHash_X11",
                "url" : "stratum+tcp://stratum.nicehash.com:3336",
                "user" : "1xxxxxxxxxxxxxxxxxxxxxxxxxxxx",
                "pass" : "x",
                "pool-algorithm" : "darkcoin-mod",
                "pool-nfactor" : "10",
                "pool-xintensity" : "300",
                "pool-thread-concurrency" : "8960",
                "pool-gpu-threads" : "2"
        },
        {
                "name" : "NiceHash_X13",
                "url" : "stratum+tcp://stratum.nicehash.com:3337",
                "user" : "1xxxxxxxxxxxxxxxxxxxxxxxxxxxx",
                "pass" : "x",
                "pool-algorithm" : "marucoin-mod",
                "pool-nfactor" : "10",
                "pool-xintensity" : "300",
                "pool-thread-concurrency" : "8192",
                "pool-gpu-threads" : "2"
        },
        {
                "name" : "NiceHash_Keccak",
...


],
"xintensity" : "4",
"vectors" : "1",
"kernel" : "zuikkis",
"lookup-gap" : "2",
"shaders" : "1280",
"thread-concurrency" : "8192",
"gpu-engine" : "1100",
"gpu-memclock" : "1500",
"gpu-fan" : "0-95",
"gpu-powertune" : "0",
"temp-cutoff" : "90",
"temp-overheat" : "85",
"temp-target" : "70",
"gpu-threads" : "2",
"auto-fan" : true,
"auto-gpu" : true,
"log" : "5",
"log-dateformat" : "1",
"failover-only" : true,
"failover-switch-delay" : "30",
"no-pool-disable" : true,
"scrypt" : true,
"queue" : "1",
"scan-time" : "7",
"expiry" : "28",
"api-listen" : true,
"api-mcast-port" : "4028",
"api-port" : "4028",
"api-allow" : "W:127.0.0.1,W:10.10......"

guzzzi

newbie

Activity: 53

Merit: 0

Is there an Problem with the NScrypt Multi-Algo Pool? When my Rig switch to another Algo (Picture to X13) they all go "Off".

another Switch

SirGeekalot

full member

Activity: 156

Merit: 100

Anyone else got any info on the Dead Multi-Algo pools?

cant get connected to any of them.

Nevermind, they came back up.

although using the zuikkis kernel gives me HW error on scrypt, switched to ckolivas, no HW errors yet but lower KHs.

Elun

member

Activity: 117

Merit: 10

Quote from: litecoinbeast on June 10, 2014, 11:10:44 PM

I can't even get this to start on windows7 x64. Starts..then closes no errors..no nothing. And trust me I have run A LOT of different cg/sgminer variations with countless kernels at this point without any problem.

After this commit: github commit opencl platfrom selecting logic changed. Try to set "gpu-platform" : "1" in conf file. You can use this exe: https://www.dropbox.com/s/fjtmzlq3o623mql/vectorAddition_exe.zip (virustotal) for check list of your platforms and devices. It's trimmed version of my opencl program for experiments. Sources, you can compile them by yourself, if don't trust me.
You may need vc++ 2010 redistributable to run it: http://www.microsoft.com/en-us/download/details.aspx?id=5555

bullus

sr. member

Activity: 294

Merit: 250

Quote from: platinum4 on June 10, 2014, 10:34:00 PM

Quote from: Elun on June 10, 2014, 08:19:20 PM

Quote from: mannyg on June 10, 2014, 03:52:17 PM

Using the new 10062014 v5 release, no matter what settings i put for gpu-threads and worksize in the conf file, they default to 2 and 256 in the .bin file name. TC & n-factor seems to be the only thing changing

Is this normal?

No, this isn't normal. Anyone who have linux, can you update your sgminer 5 and confirm that this options don't work?

Is there a Windows build for the 10th Elun?

check here
https://bitcointalksearch.org/topic/m.7226879

litecoinbeast

full member

Activity: 136

Merit: 100

I can't even get this to start on windows7 x64. Starts..then closes no errors..no nothing. And trust me I have run A LOT of different cg/sgminer variations with countless kernels at this point without any problem.

platinum4

sr. member

Activity: 547

Merit: 250

Quote from: Elun on June 10, 2014, 08:19:20 PM

Quote from: mannyg on June 10, 2014, 03:52:17 PM

Using the new 10062014 v5 release, no matter what settings i put for gpu-threads and worksize in the conf file, they default to 2 and 256 in the .bin file name. TC & n-factor seems to be the only thing changing

Is this normal?

No, this isn't normal. Anyone who have linux, can you update your sgminer 5 and confirm that this options don't work?

Is there a Windows build for the 10th Elun?

mannyg

full member

Activity: 142

Merit: 101

Quote from: Adz on June 10, 2014, 08:04:13 PM

Quote from: mannyg on June 10, 2014, 07:09:06 PM

Quote from: Adz on June 10, 2014, 06:19:16 PM

Quote from: mannyg on June 10, 2014, 05:01:06 PM

Quote from: Adz on June 10, 2014, 04:07:03 PM

Problem I have now is they are all dead.

0: Enabled Dead Quota 1 Prio 0 'NiceHash_Scrypt' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
1: Enabled Dead Quota 1 Prio 1 'NiceHash_Scrypt-N' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
2: Enabled Dead Quota 1 Prio 2 'NiceHash_X11' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
3: Enabled Dead Quota 1 Prio 3 'NiceHash_X13' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
4: Enabled Dead Quota 1 Prio 4 'NiceHash_Keccak' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
5: Enabled Dead Quota 1 Prio 5 'NiceHashScrypt_Backup' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
6: Enabled Alive Quota 1 Prio 6 'iPOMiner' User:Adzomatic.GPU1

Current pool management strategy: Failover

I can get NiceHash_Scrypt to work but only on port :3333
Any thoughts

Why is your memclock so low? 900 is pretty low even for the lowest end cards out there, you may need to raise that to something between 1200 & 1600 or so depending on your GPU

I wouldn't have thought memclock would have any baring on why the pools are all failing over? But as you asked I do have 5x low end GPUs connected (I don't have to pay power cost so running on separate machines) they are all HD 7700's. Setting the memclock to anything higher than 900 makes the rig unstable and pretty much reboots itself.

Happens constantly have tried over a few days and still no good on this front the only pool that works is the 'NiceHash_Scrypt' when I set it to port 3333

The only way I can see a low memclock setting would disable pools is if there was no work for the GPU's to do, and there is a setting somehwere to disable pools that dont have any work. If you're connecting to a pool with a high diff, you wont get much work. In your case, all the pools look dead, which when happened to me, happened because of a temp ban from too many rejected shares. Not sure if stale shares would do it also, but a slow rig on a high diff pool might have more stale shares submitted and *might* cause it to be temp banned.

As for your rig becoming unstable, its because your intensity setting is too high if you have a 7750. Set it to 11. If you have a 7770, your thread concurrency is wrong,.. set it to 8000.

Instead of me going through every possible tweak of the 7700 family cards, what model GPU do you have... I'll see if i can spot what it is in your config that is making them unstable

The Dead pools happen as soon as sgminer loads, not even enough chance to try and send anything. I have left them several hours in between attempts but same thing.

Thanks for the help mannyg. The 5x GPU's are made up of 2 separate model 7750's as below.
2x Sapphire HD 7750 2GB GDDR3 http://www.sapphiretech.com/presentation/product/?cid=1&gid=3&sgid=1159&pid=1682&lid=1
3x Gigabyte GV-R775OC-2GI http://www.gigabyte.com.au/products/product-page.aspx?pid=4345#ov

I'll give it a go setting intensity to 11 and raising the clock, let me know what you can figure out about the thread concurrency as I swear I had looked into this 6x months ago when I initially set them up and had it right...
Thanks again

Try these settings:

"intensity" : "11" OR "xintensity" : "4"
"workload" : "64"
"thread-concurrency" : "8000"
"gpu-engine" : "800-900"
"gpu-memclock" : "1125-1200"
"shaders" : "512"
"gpu-threads" : "2"
"gpu-powertune" : "-20"

Toggle the powertune to 0 and +20 and the workload to 64, 96 or 128, .. yo should eventually end up with something around 110-125KH/s on SCRYPT per GPU. You can get a higher KH but it would most likely return more HW errors or rejects.

You can also try increasing the intensity 1 or 2 more, but more than 14 wont do much. You may be better off going xintensity 4-5-6 instead

Good luck, hope it helps

Elun

member

Activity: 117

Merit: 10

Quote from: mannyg on June 10, 2014, 03:52:17 PM

Using the new 10062014 v5 release, no matter what settings i put for gpu-threads and worksize in the conf file, they default to 2 and 256 in the .bin file name. TC & n-factor seems to be the only thing changing

Is this normal?

No, this isn't normal. Anyone who have linux, can you update your sgminer 5 and confirm that this options don't work?

Adz

newbie

Activity: 24

Merit: 0

Quote from: mannyg on June 10, 2014, 07:09:06 PM

Quote from: Adz on June 10, 2014, 06:19:16 PM

Quote from: mannyg on June 10, 2014, 05:01:06 PM

Quote from: Adz on June 10, 2014, 04:07:03 PM

Problem I have now is they are all dead.

0: Enabled Dead Quota 1 Prio 0 'NiceHash_Scrypt' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
1: Enabled Dead Quota 1 Prio 1 'NiceHash_Scrypt-N' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
2: Enabled Dead Quota 1 Prio 2 'NiceHash_X11' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
3: Enabled Dead Quota 1 Prio 3 'NiceHash_X13' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
4: Enabled Dead Quota 1 Prio 4 'NiceHash_Keccak' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
5: Enabled Dead Quota 1 Prio 5 'NiceHashScrypt_Backup' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
6: Enabled Alive Quota 1 Prio 6 'iPOMiner' User:Adzomatic.GPU1

Current pool management strategy: Failover

I can get NiceHash_Scrypt to work but only on port :3333
Any thoughts

Why is your memclock so low? 900 is pretty low even for the lowest end cards out there, you may need to raise that to something between 1200 & 1600 or so depending on your GPU

I wouldn't have thought memclock would have any baring on why the pools are all failing over? But as you asked I do have 5x low end GPUs connected (I don't have to pay power cost so running on separate machines) they are all HD 7700's. Setting the memclock to anything higher than 900 makes the rig unstable and pretty much reboots itself.

Happens constantly have tried over a few days and still no good on this front the only pool that works is the 'NiceHash_Scrypt' when I set it to port 3333

The only way I can see a low memclock setting would disable pools is if there was no work for the GPU's to do, and there is a setting somehwere to disable pools that dont have any work. If you're connecting to a pool with a high diff, you wont get much work. In your case, all the pools look dead, which when happened to me, happened because of a temp ban from too many rejected shares. Not sure if stale shares would do it also, but a slow rig on a high diff pool might have more stale shares submitted and *might* cause it to be temp banned.

As for your rig becoming unstable, its because your intensity setting is too high if you have a 7750. Set it to 11. If you have a 7770, your thread concurrency is wrong,.. set it to 8000.

Instead of me going through every possible tweak of the 7700 family cards, what model GPU do you have... I'll see if i can spot what it is in your config that is making them unstable

The Dead pools happen as soon as sgminer loads, not even enough chance to try and send anything. I have left them several hours in between attempts but same thing.

Thanks for the help mannyg. The 5x GPU's are made up of 2 separate model 7750's as below.
2x Sapphire HD 7750 2GB GDDR3 http://www.sapphiretech.com/presentation/product/?cid=1&gid=3&sgid=1159&pid=1682&lid=1
3x Gigabyte GV-R775OC-2GI http://www.gigabyte.com.au/products/product-page.aspx?pid=4345#ov

I'll give it a go setting intensity to 11 and raising the clock, let me know what you can figure out about the thread concurrency as I swear I had looked into this 6x months ago when I initially set them up and had it right...
Thanks again

mannyg

full member

Activity: 142

Merit: 101

Quote from: Adz on June 10, 2014, 06:19:16 PM

Quote from: mannyg on June 10, 2014, 05:01:06 PM

Quote from: Adz on June 10, 2014, 04:07:03 PM

Problem I have now is they are all dead.

0: Enabled Dead Quota 1 Prio 0 'NiceHash_Scrypt' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
1: Enabled Dead Quota 1 Prio 1 'NiceHash_Scrypt-N' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
2: Enabled Dead Quota 1 Prio 2 'NiceHash_X11' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
3: Enabled Dead Quota 1 Prio 3 'NiceHash_X13' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
4: Enabled Dead Quota 1 Prio 4 'NiceHash_Keccak' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
5: Enabled Dead Quota 1 Prio 5 'NiceHashScrypt_Backup' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
6: Enabled Alive Quota 1 Prio 6 'iPOMiner' User:Adzomatic.GPU1

Current pool management strategy: Failover

I can get NiceHash_Scrypt to work but only on port :3333
Any thoughts

Why is your memclock so low? 900 is pretty low even for the lowest end cards out there, you may need to raise that to something between 1200 & 1600 or so depending on your GPU

I wouldn't have thought memclock would have any baring on why the pools are all failing over? But as you asked I do have 5x low end GPUs connected (I don't have to pay power cost so running on separate machines) they are all HD 7700's. Setting the memclock to anything higher than 900 makes the rig unstable and pretty much reboots itself.

Happens constantly have tried over a few days and still no good on this front the only pool that works is the 'NiceHash_Scrypt' when I set it to port 3333

The only way I can see a low memclock setting would disable pools is if there was no work for the GPU's to do, and there is a setting somehwere to disable pools that dont have any work. If you're connecting to a pool with a high diff, you wont get much work. In your case, all the pools look dead, which when happened to me, happened because of a temp ban from too many rejected shares. Not sure if stale shares would do it also, but a slow rig on a high diff pool might have more stale shares submitted and *might* cause it to be temp banned.

As for your rig becoming unstable, its because your intensity setting is too high if you have a 7750. Set it to 11. If you have a 7770, your thread concurrency is wrong,.. set it to 8000.

Instead of me going through every possible tweak of the 7700 family cards, what model GPU do you have... I'll see if i can spot what it is in your config that is making them unstable

Adz

newbie

Activity: 24

Merit: 0

Quote from: mannyg on June 10, 2014, 05:01:06 PM

Quote from: Adz on June 10, 2014, 04:07:03 PM

Problem I have now is they are all dead.

0: Enabled Dead Quota 1 Prio 0 'NiceHash_Scrypt' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
1: Enabled Dead Quota 1 Prio 1 'NiceHash_Scrypt-N' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
2: Enabled Dead Quota 1 Prio 2 'NiceHash_X11' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
3: Enabled Dead Quota 1 Prio 3 'NiceHash_X13' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
4: Enabled Dead Quota 1 Prio 4 'NiceHash_Keccak' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
5: Enabled Dead Quota 1 Prio 5 'NiceHashScrypt_Backup' User:1NHDHBMF7R3F4RtCFpA1mzNBnWCAjNyC9er
6: Enabled Alive Quota 1 Prio 6 'iPOMiner' User:Adzomatic.GPU1

Current pool management strategy: Failover

I can get NiceHash_Scrypt to work but only on port :3333
Any thoughts

Why is your memclock so low? 900 is pretty low even for the lowest end cards out there, you may need to raise that to something between 1200 & 1600 or so depending on your GPU

I wouldn't have thought memclock would have any baring on why the pools are all failing over? But as you asked I do have 5x low end GPUs connected (I don't have to pay power cost so running on separate machines) they are all HD 7700's. Setting the memclock to anything higher than 900 makes the rig unstable and pretty much reboots itself.

Happens constantly have tried over a few days and still no good on this front the only pool that works is the 'NiceHash_Scrypt' when I set it to port 3333

W0rkH0rse

member

Activity: 80

Merit: 10

Quote from: lewisc46 on June 10, 2014, 01:22:45 AM

I'm using BAMT 1.6.9 and I've compiled sgminer amd updated the bamt.conf with my pool details but whenever I restart the mining i just get a load of invalids.

Bad news: there is no BAMT 1.6.9 version.

Good news: your botnet zombie is probably working perfectly.

MysteryX

member

Activity: 98

Merit: 10

4xxx ports aren't working for me, they go straight to my failover pool, but mining on the 3xxx ports by themselves are working Huh

mannyg

full member

Activity: 142

Merit: 101

cgladue

member

Activity: 97

Merit: 10

Topic: [ANN] sgminer v5 - optimized X11/X13/NeoScrypt/Lyra2RE/etc. kernel-switch miner - page 236. (Read 877859 times)