Pages:
Author

Topic: HSRMINER Neoscrypt Fork by Justaminer - High Hashrate, API, all GPUs supported! - page 11. (Read 13383 times)

jr. member
Activity: 325
Merit: 2
Testing last version from 25.02.2018

INFO : [14:23:16] : GPU#0 - ASUS GTX 1070, speed is 829.04kH/s
INFO : [14:23:16] : GPU#4 - Gigabyte GTX 980, speed is 696.24kH/s
INFO : [14:23:16] : GPU#7 - Gigabyte GTX 1070, speed is 1108.09kH/s
INFO : [14:23:16] : GPU#8 - Zotac GTX 980 Ti, speed is 820.14kH/s
INFO : [14:23:16] : GPU#9 - Gigabyte GTX 1070, speed is 1110.91kH/s

INFO : [14:38:15] : Share Accepted, overall speed is 4541.39kH/s, efficiency 100.00%

70-80 PL, +100 Core, +400 Memory

Windows 10 v1703
Video Driver: 390.65
GPUS:
Asus GTX 1070 - Expedition
2x Gigabyte GTX 1070 - Gaming 1
Gigabyte GTX 980 - Windforce 3x
Zotac GTX 980Ti - AMP! Edition

Tested with default intensity and increased priority version. I can benchmark full rig from 11 cards - 4x1050Ti, 2x1060-6Gb, 3x1070, 1x980, 1x980Ti, if anybody want it.

Thanks for the info, what's the problem with GPU#0 - ASUS GTX 1070, speed is 829.04kH/s ? Is it downclocked or you've been using it during benchmark?
jr. member
Activity: 325
Merit: 2
Still crashing for me.

Do you have Titan XP? It will be supported with the next release.
jr. member
Activity: 325
Merit: 2
Are you going to let us make changes to the intensity?  I see some of my cards tops out at 16.5.

-i 7 = 16.5 Intensity

OR are you using this as what say, "-i 7 = use as much of the VRAM as possible"

All in all, I like what you and your team has done.  Thanks to you all my GPUs are full and digging away.  Can't wait to see what else you can do.

HSR is also supported by this software?

Current version doesn't allow to change intensity with --benchmark, i.e. even if you specify -i N, it will be reset to defaults anyway before benchmark - it's default behavior of hsrminer. I will change that with the next release, so you will be able to actually change it and see effect with --benchmark.

Default intensity for 9xx GPU right now is 16.1, I've tried to change it, doesn't have much effect, but you can try anyway. For 970 you have to watch out not to cross 3.5 gb memory limit, or GPU will hash very very slow, use tools like GPU-Z to see video memory usage.

There is no team, I'm doing all the work myself. This is neoscrypt only software.
jr. member
Activity: 325
Merit: 2
Windows 10 1709 (build 16299.248)
3x nVidia 1070 Ti @ 100% PL, +200 core, +700 memory
Biostar 250-BTC
4gb ram
80gb page file

With 388 this worked just fine with no errors.

Jumping to 390 this causes errors. Even at 70% PL same thing. I had to drop it down to 50% and only then it would not cause that error. So I downgraded back to 388 drivers and run at 100% PL right now until we can debug it?

I don't know if I will be able to debug such problem. Don't you think that +200 core + 700 mem is too much for your gpu? Have you tried to reduce overclocking with 390.77 drivers?
newbie
Activity: 3
Merit: 0
Testing last version from 25.02.2018

INFO : [14:23:16] : GPU#0 - ASUS GTX 1070, speed is 829.04kH/s
INFO : [14:23:16] : GPU#4 - Gigabyte GTX 980, speed is 696.24kH/s
INFO : [14:23:16] : GPU#7 - Gigabyte GTX 1070, speed is 1108.09kH/s
INFO : [14:23:16] : GPU#8 - Zotac GTX 980 Ti, speed is 820.14kH/s
INFO : [14:23:16] : GPU#9 - Gigabyte GTX 1070, speed is 1110.91kH/s

INFO : [14:38:15] : Share Accepted, overall speed is 4541.39kH/s, efficiency 100.00%

70-80 PL, +100 Core, +400 Memory

Windows 10 v1703
Video Driver: 390.65
GPUS:
Asus GTX 1070 - Expedition
2x Gigabyte GTX 1070 - Gaming 1
Gigabyte GTX 980 - Windforce 3x
Zotac GTX 980Ti - AMP! Edition

Tested with default intensity and increased priority version. I can benchmark full rig from 11 cards - 4x1050Ti, 2x1060-6Gb, 3x1070, 1x980, 1x980Ti, if anybody want it.
newbie
Activity: 7
Merit: 0
Still crashing for me.

Code:
hsrminer_neoscrypt_fork.exe -a neoscrypt -o stratum+tcp://hub.miningpoolhub.com:17012 -u Zackdaystar.miner1 -p x


              888888ba           oo       dP oo
              88    `8b                   88
              88     88 dP   .dP dP .d888b88 dP .d8888b.
              88     88 88   d8' 88 88'  `88 88 88'  `88     888888b                   8P   dP
              88     88 88 .88'  88 88.  .88 88 88.  .88     88      .d8888b. 88d888b. 88  d8
              dP     dP 8888P'   dP `88888P8 dP `88888P8     888888  88'  `88 88'  `88 88d88
                                                             88      88.  .88 88       88  d8
                                                             dP      `88888P' dP       8P   dP
  dP                                    oo
  88
  88d888b. .d8888b. 88d888b. 88d8b.d8b. dP 88d888b. .d8888b. 88d888b.
  88'  `88 Y8ooooo. 88'  `88 88'`88'`88 88 88'  `88 88ooood8 88'  `88
  88    88       88 88       88  88  88 88 88    88 88.  ... 88
  dP    dP `88888P' dP       dP  dP  dP dP dP    dP `88888P' dP

                                                             https://github.com/justaminer/hsrm-fork
****************************************************************************************************
 NVIDIA hsrminer 1.0.1+ for Neoscrypt forked of palgin by Justaminer.    Fork's version: 25.02.2018
****************************************************************************************************
 Fork contains 0% devfee, working API and -r option, enjoy! ( run with -h to see all new options! )
****************************************************************************************************

 INFO : [10:31:23] : POOL: stratum+tcp://hub.miningpoolhub.com:17012
 INFO : [10:31:23] : Checking for NVML...
 INFO : [10:31:23] : NVML monitoring enabled, continue...
 INFO : [10:31:23] : 1 GPU found/specified, initializing miner...
 INFO : [10:31:24] : Share difficulty reset -- 32 (0.00049)

and this in the event logs
Code:
-
-
 
  1000
  2
  100
  0x80000000000000
 
  59361
  Application
  black-gold
 
 

-
  hsrminer_neoscrypt.exe
  1.0.1.0
  5a47af32
  hsrminer_neoscrypt.exe
  1.0.1.0
  5a47af32
  c0000005
  000000000012e349
  2248
  01d3ae228741f3ac
  C:\MultiPoolMinerV2\Bin\NeoScrypt-Palgin\hsrminer_neoscrypt.exe
  C:\MultiPoolMinerV2\Bin\NeoScrypt-Palgin\hsrminer_neoscrypt.exe
  88576d72-a9a0-40c7-9172-385945a73237
 
 
 

 





member
Activity: 104
Merit: 18
Are you going to let us make changes to the intensity?  I see some of my cards tops out at 16.5.

-i 7 = 16.5 Intensity

OR are you using this as what say, "-i 7 = use as much of the VRAM as possible"

All in all, I like what you and your team has done.  Thanks to you all my GPUs are full and digging away.  Can't wait to see what else you can do.

HSR is also supported by this software?
newbie
Activity: 20
Merit: 0
Windows 10 1709 (build 16299.248)
3x nVidia 1070 Ti @ 100% PL, +200 core, +700 memory
Biostar 250-BTC
4gb ram
80gb page file

With 388 this worked just fine with no errors.

Jumping to 390 this causes errors. Even at 70% PL same thing. I had to drop it down to 50% and only then it would not cause that error. So I downgraded back to 388 drivers and run at 100% PL right now until we can debug it?
jr. member
Activity: 325
Merit: 2
For anyone having the miner stop with the following error:

Quote
Cuda error in func 'neoscrypt_cpu_hash_1070' at line 1460

It's because of newer nVidia drivers. Can confirm it happened to me on latest drivers (390.x) and when I downgraded back to 388.3x it seems to work again. Looks to be an issue with original palgin miner too as per: https://github.com/palginpav/hsrminer/issues/9

Hopefully this fork can be updated so we can use latest drivers in case of any future speed ups happen.

I need more info, OS, hardware? Because I don't have any problems with GTX 1070 and 390.77 driver on Win 10 v1709.
member
Activity: 104
Merit: 18
For anyone having the miner stop with the following error:

Quote
Cuda error in func 'neoscrypt_cpu_hash_1070' at line 1460

It's because of newer nVidia drivers. Can confirm it happened to me on latest drivers (390.x) and when I downgraded back to 388.3x it seems to work again. Looks to be an issue with original palgin miner too as per: https://github.com/palginpav/hsrminer/issues/9

Hopefully this fork can be updated so we can use latest drivers in case of any future speed ups happen.

Before his update, I'd get that with my 980's
newbie
Activity: 45
Merit: 0
p104-100 1100kh/s
pl 90%
core +100
mem +400

Thanks for the info. Have you tested it as single gpu only? Can you benchmark 2+ gpu simultaneously to be sure that everything is ok?

2 gpu rig and 8 gpu rig , both do 1100kh/s / card ( 1090 -> 1115 kh/s)
newbie
Activity: 20
Merit: 0
For anyone having the miner stop with the following error:

Quote
Cuda error in func 'neoscrypt_cpu_hash_1070' at line 1460

It's because of newer nVidia drivers. Can confirm it happened to me on latest drivers (390.x) and when I downgraded back to 388.3x it seems to work again. Looks to be an issue with original palgin miner too as per: https://github.com/palginpav/hsrminer/issues/9

Hopefully this fork can be updated so we can use latest drivers in case of any future speed ups happen.
jr. member
Activity: 325
Merit: 2
It works with 1080TI. It works with 980TI. But isn't working with 1080ti+980ti at same time! How can I solve it? Great work

Try latest version, should work now.
jr. member
Activity: 325
Merit: 2
Are you planning to make monitoring of the program through a browser like this done in dstm's?

API is ccminer-compatible, and ccminer doesn't have such feature. So may be in the future, this is not a priority task right now.
jr. member
Activity: 325
Merit: 2
p104-100 1100kh/s
pl 90%
core +100
mem +400

Thanks for the info. Have you tested it as single gpu only? Can you benchmark 2+ gpu simultaneously to be sure that everything is ok?
jr. member
Activity: 325
Merit: 2
INFO : [12:34:26] : Starting benchmark mode with neoscrypt
 INFO : [12:34:26] : Checking for NVML...
 INFO : [12:34:26] : NVML monitoring enabled, continue...
 INFO : [12:34:26] : 2 GPUs found/specified, initializing miner...
 INFO : [12:34:27] : GPU #3: GeForce GTX 980, flags: 0, 0, 1
 INFO : [12:34:27] : GPU #7: GeForce GTX 980, flags: 0, 0, 1
 INFO : [12:34:35] : GPU #7 Found nonce 804cc3e4
 INFO : [12:34:43] : GPU #7 Found nonce 804cc3e4
 INFO : [12:34:43] : GPU #3 Found nonce 009cb91b
 INFO : [12:34:59] : GPU #3 Found nonce 009cb91b
 INFO : [12:35:00] : GPU#7 - neoscrypt hashrate = 682.04kH/s
 INFO : [12:35:00] : GPU#3 - neoscrypt hashrate = 677.73kH/s
 INFO : [12:35:01] : Result for GPU#3 - GeForce GTX 980: 677.73 kH/s
 INFO : [12:35:01] : Result for GPU#7 - GeForce GTX 980: 682.04 kH/s
 INFO : [12:35:01] : Total: 1359.76 kH/s

both are set at:
tdp=100%
core=+100 / 1354
memory=+100 / 3704
vram size per card is 4096mb

Let me know when you want the 960 tested out.  I've made an nice swap point if need be


Ok, thanks for testing. Support for 960 will added with the next release.
newbie
Activity: 45
Merit: 0
p104-100 1100kh/s
pl 90%
core +100
mem +400
member
Activity: 104
Merit: 18
 INFO : [12:34:26] : Starting benchmark mode with neoscrypt
 INFO : [12:34:26] : Checking for NVML...
 INFO : [12:34:26] : NVML monitoring enabled, continue...
 INFO : [12:34:26] : 2 GPUs found/specified, initializing miner...
 INFO : [12:34:27] : GPU #3: GeForce GTX 980, flags: 0, 0, 1
 INFO : [12:34:27] : GPU #7: GeForce GTX 980, flags: 0, 0, 1
 INFO : [12:34:35] : GPU #7 Found nonce 804cc3e4
 INFO : [12:34:43] : GPU #7 Found nonce 804cc3e4
 INFO : [12:34:43] : GPU #3 Found nonce 009cb91b
 INFO : [12:34:59] : GPU #3 Found nonce 009cb91b
 INFO : [12:35:00] : GPU#7 - neoscrypt hashrate = 682.04kH/s
 INFO : [12:35:00] : GPU#3 - neoscrypt hashrate = 677.73kH/s
 INFO : [12:35:01] : Result for GPU#3 - GeForce GTX 980: 677.73 kH/s
 INFO : [12:35:01] : Result for GPU#7 - GeForce GTX 980: 682.04 kH/s
 INFO : [12:35:01] : Total: 1359.76 kH/s

both are set at:
tdp=100%
core=+100 / 1354
memory=+100 / 3704
vram size per card is 4096mb

Let me know when you want the 960 tested out.  I've made an nice swap point if need be
GVG
newbie
Activity: 2
Merit: 0
Are you planning to make monitoring of the program through a browser like this done in dstm's?
jr. member
Activity: 325
Merit: 2
I *think* this miner BSODed my rig this morning with a Windows error code of 0x...50 which is usually from a hardware failure, but seeing as this miner now warns it will stress the GPUs even harder than its progenitor (which is already quite finicky about overclocking) I'm guess it is the actual culprit.

Should I tweak intensity, or go with the non-hp version?


Try non-hp version.
Pages:
Jump to: