Open Source XPM (Primecoin) GPU Miner & Pool xpmforall.org

[Pinocchio]

What CPD can we expect from Radeon VII, and is there going to be any work done on optimizing this miner for it as well?

[eXtremal.ik71]

Could you probably add altilly.com exchanger to xpmforall.org exchanges list?

Volume 0.072 BTC, not so much.

(Sometimes after creating new bin file miner closes randomly but you can just re-start it).

Possibly fixed in xpmclient 10.3

What CPD can we expect from Radeon VII, and is there going to be any work done on optimizing this miner for it as well?

Can't predict, need test Radeon VII.. and RTX2080Ti in new version.

[tar1s]

Vega -> +0, 03 (0.04) cpd
390 and 480 -> -0.03 (0.04) cpd

[eXtremal.ik71]

Vega -> +0, 03 (0.04) cpd
390 and 480 -> -0.03 (0.04) cpd

Vega56 must show ~0.7 CPD now

Miner screenshot from Windows 10 (stock Vega56, Adrenalin 19.3.2, power limit set to 30% for stabilize core clock):
http://i66.tinypic.com/k3p3m8.png

2019-03-17 16:47:29.258 (1257.849s) INFO| Work received: height=3097070 diff=11.091829
2019-03-17 16:48:14.756 (1303.347s) INFO| [GPU 0] T=81C A=97% E=0 primes=0.103879 fermat=544908/sec cpd=0.72/day
2019-03-17 16:48:14.761 (1303.353s) INFO| GPU 0: core=1466MHz mem=800MHz powertune=30 fanspeed=67
2019-03-17 16:48:14.767 (1303.358s) INFO| (ST/INV/DUP): 63x 7ch(13/0/0) 9x 8ch(3/0/0) 2x 9ch(1/0/0)
2019-03-17 16:49:01.256 (1349.848s) WARN| Stale share 200ms
2019-03-17 16:49:01.458 (1350.049s) INFO| Work received: height=3097072 diff=11.091718
2019-03-17 16:49:01.470 (1350.061s) WARN| ran out of hashes, increasing sha256 work size coefficient to 131072
2019-03-17 16:49:14.615 (1363.206s) INFO| [GPU 0] T=81C A=97% E=0 primes=0.103870 fermat=543283/sec cpd=0.71/day
2019-03-17 16:49:14.620 (1363.211s) INFO| (ST/INV/DUP): 66x 7ch(14/0/0) 9x 8ch(3/0/0) 2x 9ch(1/0/0)
2019-03-17 16:49:19.163 (1367.757s) WARN| Stale share 201ms
2019-03-17 16:49:19.367 (1367.958s) INFO| Work received: height=3097073 diff=11.091871
2019-03-17 16:49:54.900 (1403.493s) WARN| Stale share 201ms
2019-03-17 16:49:55.103 (1403.694s) INFO| Work received: height=3097074 diff=11.091998
2019-03-17 16:50:14.455 (1423.046s) INFO| [GPU 0] T=81C A=98% E=0 primes=0.103902 fermat=534350/sec cpd=0.70/day
2019-03-17 16:50:14.460 (1423.052s) INFO| (ST/INV/DUP): 68x 7ch(16/0/0) 9x 8ch(3/0/0) 2x 9ch(1/0/0)
2019-03-17 16:50:14.856 (1423.455s) INFO| Work received: height=3097075 diff=11.092142

Average is 0.71 CPD, 10.2.2 shows 0.6-0.62 CPD in same configuration.

Config:

cpuload = "1";
target = "auto";
sieveSize = "480";
weaveDepth = "36864";
width = "auto";
windowSize = "8192";
multiplierLimits = ["25", "31", "34"];

Benchmarks result from xpmclient -b (look like driver not best for vega's):

2019-03-17 16:50:49.131 (   9.045s) INFO| square 320 bits: 36.531ms (3674.078M ops/sec)
2019-03-17 16:50:52.662 (  12.575s) INFO| multiply 320 bits: 31.002ms (4329.325M ops/sec)
2019-03-17 16:50:56.621 (  16.530s) INFO| square 352 bits: 46.268ms (2900.876M ops/sec)
2019-03-17 16:51:01.429 (  21.347s) INFO| multiply 352 bits: 37.778ms (3552.801M ops/sec)
2019-03-17 16:51:03.677 (  23.596s) INFO| Fermat tests 320 bits: 29.894ms (4.385M ops/sec, single thread cpu: 0.064M ops/sec)
2019-03-17 16:51:06.678 (  26.592s) INFO| Fermat tests 352 bits: 39.578ms (3.312M ops/sec, single thread cpu: 0.046M ops/sec)
2019-03-17 16:51:06.678 (  26.601s) INFO| *** hashmod benchmark ***
2019-03-17 16:51:07.148 (  27.059s) INFO|  MHash per second: 1281.229
2019-03-17 16:51:07.148 (  27.064s) INFO|  Hash per iteration: 10.781 (0.000129 %)
2019-03-17 16:51:07.148 (  27.069s) INFO|  Average hash multiplier size: 28.216
2019-03-17 16:51:07.163 (  27.076s) INFO|    Hashed with primorial 13 is 22.464%
2019-03-17 16:51:07.171 (  27.081s) INFO|    Hashed with primorial 14 is 62.464%
2019-03-17 16:51:07.178 (  27.087s) INFO|    Hashed with primorial 15 is 15.072%
2019-03-17 16:51:07.179 (  27.093s) INFO| *** sieve (check) benchmark ***
2019-03-17 16:51:15.071 (  34.986s) INFO|  * [OK] found candidates by CPU: 7218 by GPU: 7228
2019-03-17 16:51:15.071 (  34.991s) INFO|  * [OK] invalid candidates: 0
2019-03-17 16:51:15.087 (  34.997s) INFO|  * [OK] CPU/GPU candidates difference: 0
2019-03-17 16:51:15.133 (  35.058s) INFO| *** sieve (performance) benchmark ***
2019-03-17 16:51:15.741 (  35.664s) INFO|  * scan speed: 147.025 G
2019-03-17 16:51:15.757 (  35.668s) INFO|  * iteration time: 9.414ms
2019-03-17 16:51:15.757 (  35.673s) INFO|  * candidates per second: 614598.160
2019-03-17 16:51:15.757 (  35.678s) INFO|  * candidates per iteration: 5785.95 (2690.40 320bit, 3095.55 352bit)
2019-03-17 16:51:15.773 (  35.686s) INFO|  * 320bit/352bit ratio: 0.869/1

[tar1s]

Vega56 must show ~0.7 CPD now

0.74(75),
 it was 0.79, but I lost the settings
580-> 0,45(46)

[pepi]

Could you probably add altilly.com exchanger to xpmforall.org exchanges list?

Volume 0.072 BTC, not so much.

(Sometimes after creating new bin file miner closes randomly but you can just re-start it).

Possibly fixed in xpmclient 10.3

What CPD can we expect from Radeon VII, and is there going to be any work done on optimizing this miner for it as well?

Can't predict, need test Radeon VII.. and RTX2080Ti in new version.

Solo mining with this client yes/no?

[Wolf9466]

Vega -> +0, 03 (0.04) cpd
390 and 480 -> -0.03 (0.04) cpd

Vega56 must show ~0.7 CPD now

Miner screenshot from Windows 10 (stock Vega56, Adrenalin 19.3.2, power limit set to 30% for stabilize core clock):
http://i66.tinypic.com/k3p3m8.png

2019-03-17 16:47:29.258 (1257.849s) INFO| Work received: height=3097070 diff=11.091829
2019-03-17 16:48:14.756 (1303.347s) INFO| [GPU 0] T=81C A=97% E=0 primes=0.103879 fermat=544908/sec cpd=0.72/day
2019-03-17 16:48:14.761 (1303.353s) INFO| GPU 0: core=1466MHz mem=800MHz powertune=30 fanspeed=67
2019-03-17 16:48:14.767 (1303.358s) INFO| (ST/INV/DUP): 63x 7ch(13/0/0) 9x 8ch(3/0/0) 2x 9ch(1/0/0)
2019-03-17 16:49:01.256 (1349.848s) WARN| Stale share 200ms
2019-03-17 16:49:01.458 (1350.049s) INFO| Work received: height=3097072 diff=11.091718
2019-03-17 16:49:01.470 (1350.061s) WARN| ran out of hashes, increasing sha256 work size coefficient to 131072
2019-03-17 16:49:14.615 (1363.206s) INFO| [GPU 0] T=81C A=97% E=0 primes=0.103870 fermat=543283/sec cpd=0.71/day
2019-03-17 16:49:14.620 (1363.211s) INFO| (ST/INV/DUP): 66x 7ch(14/0/0) 9x 8ch(3/0/0) 2x 9ch(1/0/0)
2019-03-17 16:49:19.163 (1367.757s) WARN| Stale share 201ms
2019-03-17 16:49:19.367 (1367.958s) INFO| Work received: height=3097073 diff=11.091871
2019-03-17 16:49:54.900 (1403.493s) WARN| Stale share 201ms
2019-03-17 16:49:55.103 (1403.694s) INFO| Work received: height=3097074 diff=11.091998
2019-03-17 16:50:14.455 (1423.046s) INFO| [GPU 0] T=81C A=98% E=0 primes=0.103902 fermat=534350/sec cpd=0.70/day
2019-03-17 16:50:14.460 (1423.052s) INFO| (ST/INV/DUP): 68x 7ch(16/0/0) 9x 8ch(3/0/0) 2x 9ch(1/0/0)
2019-03-17 16:50:14.856 (1423.455s) INFO| Work received: height=3097075 diff=11.092142

Average is 0.71 CPD, 10.2.2 shows 0.6-0.62 CPD in same configuration.

Config:

cpuload = "1";
target = "auto";
sieveSize = "480";
weaveDepth = "36864";
width = "auto";
windowSize = "8192";
multiplierLimits = ["25", "31", "34"];

Benchmarks result from xpmclient -b (look like driver not best for vega's):

2019-03-17 16:50:49.131 (   9.045s) INFO| square 320 bits: 36.531ms (3674.078M ops/sec)
2019-03-17 16:50:52.662 (  12.575s) INFO| multiply 320 bits: 31.002ms (4329.325M ops/sec)
2019-03-17 16:50:56.621 (  16.530s) INFO| square 352 bits: 46.268ms (2900.876M ops/sec)
2019-03-17 16:51:01.429 (  21.347s) INFO| multiply 352 bits: 37.778ms (3552.801M ops/sec)
2019-03-17 16:51:03.677 (  23.596s) INFO| Fermat tests 320 bits: 29.894ms (4.385M ops/sec, single thread cpu: 0.064M ops/sec)
2019-03-17 16:51:06.678 (  26.592s) INFO| Fermat tests 352 bits: 39.578ms (3.312M ops/sec, single thread cpu: 0.046M ops/sec)
2019-03-17 16:51:06.678 (  26.601s) INFO| *** hashmod benchmark ***
2019-03-17 16:51:07.148 (  27.059s) INFO|  MHash per second: 1281.229
2019-03-17 16:51:07.148 (  27.064s) INFO|  Hash per iteration: 10.781 (0.000129 %)
2019-03-17 16:51:07.148 (  27.069s) INFO|  Average hash multiplier size: 28.216
2019-03-17 16:51:07.163 (  27.076s) INFO|    Hashed with primorial 13 is 22.464%
2019-03-17 16:51:07.171 (  27.081s) INFO|    Hashed with primorial 14 is 62.464%
2019-03-17 16:51:07.178 (  27.087s) INFO|    Hashed with primorial 15 is 15.072%
2019-03-17 16:51:07.179 (  27.093s) INFO| *** sieve (check) benchmark ***
2019-03-17 16:51:15.071 (  34.986s) INFO|  * [OK] found candidates by CPU: 7218 by GPU: 7228
2019-03-17 16:51:15.071 (  34.991s) INFO|  * [OK] invalid candidates: 0
2019-03-17 16:51:15.087 (  34.997s) INFO|  * [OK] CPU/GPU candidates difference: 0
2019-03-17 16:51:15.133 (  35.058s) INFO| *** sieve (performance) benchmark ***
2019-03-17 16:51:15.741 (  35.664s) INFO|  * scan speed: 147.025 G
2019-03-17 16:51:15.757 (  35.668s) INFO|  * iteration time: 9.414ms
2019-03-17 16:51:15.757 (  35.673s) INFO|  * candidates per second: 614598.160
2019-03-17 16:51:15.757 (  35.678s) INFO|  * candidates per iteration: 5785.95 (2690.40 320bit, 3095.55 352bit)
2019-03-17 16:51:15.773 (  35.686s) INFO|  * 320bit/352bit ratio: 0.869/1

Is this still open source?

[suona]

390 and 480 -> -0.03 (0.04) cpd

Same decrease on R9 380 and R9 390. Driver version is 19.3.2.

[burnwaygta4]

390 and 480 -> -0.03 (0.04) cpd

Same decrease on R9 380 and R9 390. Driver version is 19.3.2.

what you R9 390 cpd ?

[eXtremal.ik71]

0.74(75),
 it was 0.79, but I lost the settings
580-> 0,45(46)

Known problem, some drivers create non-optimal kernels.. for example, kernel compiled for R9 29x on old drivers works better on Polaris GPUs. Need fix it in next miner version 10.4.

Пpивeтcтвyю, в чём x64 вepcия xpmclient "лyчшe" x86 ?
Бoльшe 2гб oзy для пpoцecca xpmclient вpoдe нe нyжнo, нaгpyзкa нa пpoцeccop низкaя дaжe в x86 вepcии. A вoт kernel's мeждy x86 и x64 вepcиями нe coвмecтимы.
Пpocьбa coбepитe x86 вepcию для xpmclient 10.3

Need use English here.
Miner users reported error "64 bit kernels not supported on 32 bit executable" in some cases, x86_64 version not have this issue.
I can add x86 build support to deploy system (https://github.com/eXtremal-ik7/xpmclient/tree/version/zcash/contrib), after this you can build x86 version by yourself.

PS нaпиcaл бы в личкy, дa y вac зaкpытo

I need use new newbie account, forum have not technical support, administrators not answer.

Solo mining with this client yes/no?

For solo mining need pool build with dropped accounting, web and other unnecessary parts. Also, pool required patched primecoind client and linux only. Need rewrite some code of pool, I working on it.
Why you not want use madmax's client for solo mining?
https://github.com/madMAx43v3r/xpmserver

Is this still open source?

xpmclient alltimes open source
https://github.com/eXtremal-ik7/xpmclient/tree/version/zcash

Same decrease on R9 380 and R9 390. Driver version is 19.3.2.

I have checked R9 390, but used little bit different config:

sieveSize = "630";
multiplierLimits = ["25", "31", "34"];

And got same results as 10.2.2 miner.

[pepi]

Quote from: pepi on March 17, 2019, 12:16:39 PM
Solo mining with this client yes/no?
For solo mining need pool build with dropped accounting, web and other unnecessary parts. Also, pool required patched primecoind client and linux only. Need rewrite some code of pool, I working on it.
Why you not want use madmax's client for solo mining?
https://github.com/madMAx43v3r/xpmserver

1. Because it is 5 years old
2. Dont know hot to compile it
3. Datacoin mine solo without any problem,so only problem we have is primecoin wallet. And Primecoin wallet is one of few that native doesnot support solo mining ( and long time ago it was promised that will be fixed but we are still waiting for that)
4. That old code has some issues with changing IP address so I dont wont to mine many hours for nothing.
Internal changes are rewritten to code to fix that  but they was never released to github 

[FablomsFDP]

Is the speed the same?

[FablomsFDP]

I did not notice any changes in speed for 580 - as it was, 0.44 remained.

[buny]

I did not notice any changes in speed for 580 - as it was, 0.44 remained.

similarly.

[shek_spir]

Hello!
I have a problem
Earlier I used the R390 + Radeon Pro Duo. Connection was via R390
Then R390 burned down due to a failed fan))
Now Pro Duo can not run for mining. The client starts for 1 second and immediately closes itself.
A similar situation on the second computer with R5 2400G
What to do?))

that's what he writes in the logs
2019-03-21 14:33:41.043 (   0.196s)  ERR| OpenCL error: -42 at /home/user/build/xpmclient/src/opencl.cpp:142

[tar1s]

what you R9 390 cpd ?

18.8.1
10.2.2->~0.47
10.3->~0.44

[drozenski]

Lost quite a bit of speed with the new version..

Testing with datacoin

10.2.2 > 178-182 CPD on all 3 of my cards. 2x RX580 and 1 R9 280

10. 3 > 160 - 165 CPD on each card if i play with the settings i can get the RX580's to 170-175 CPD but the R9280 Still sits in the low 160's

[burnwaygta4]

what you R9 390 cpd ?

18.8.1
10.2.2->~0.47
10.3->~0.44

core frequency ?

[tar1s]

core frequency ?

[wsr22sdk]

Hi guys,
I have this problem with the new client:

• GeForce GTX 1080 Ti; Compute capability 6.1

<info> compiling ...
<info> source: 222565 bytes
xpm.cu(6885): error: identifier "SIEVERANGE3" is undefined

xpm.cu(6896): error: identifier "SIEVERANGE1" is undefined

xpm.cu(6910): error: identifier "SIEVERANGE2" is undefined

xpm.cu(6954): error: identifier "SIEVERANGE3" is undefined

4 errors detected in the compilation of "xpm.cu".