Bitcoin Forum
June 25, 2024, 12:19:14 PM *
News: Latest Bitcoin Core release: 27.0 [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 »
  Print  
Author Topic: hsrminer - Nvidia mining software for various algos by palgin&alexkap  (Read 30738 times)
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 25, 2017, 01:05:44 PM
 #221

So, update i have been using since last 4 hours and i see that its stable 99.67% efficiency. im getting with one 1080ti for neoscrypt 1530.33 Khs on avg. overall i get 10% more speed than CCMiner which is great.

im running at 90% power limit
no memory overclock
+50 core overclock

min power usage as per software: 179W Avg: 197w Max: 209w

speed 1530 Khs.

I see you have mentioned in first post as 1850 Khs on 1080ti is there somethin i can tweak. Also i see memory is fully utilized 10625MB is this ok?



As marked in the first post, it's user-reported speed:

Was going to update previous post but there were more replies to the thread than I expected.

Environment is inside closed cases ambient temps are 72F

Neoscrypt hashrates using Klaust
Getting 1434KH/s Temp 70c on stock clocks 1080TI FE
Getting 1186KH/s Temp 78c using stock clocks on my Zotac Mini 1080

HSRminer hashrates:
1080TI FE
No adjustments 26.9% increase miner side
speed is 1933.92kH/s, Temp 69c, zPool.ca 45 Min average shows 2133KH/s

Zotac Mini 1080
No adjustments 8.86% loss miner side
speed is  1016.32kH/s, Temp 80c, zpool.ca 45 Min average shows 1216Kh/s

guessing miner hash is calculating after devfee adjustment.

Now I just need to be able to fetch ccminer style stats and I'm using it for my 1080TI

1070: same speed as Klaust 8.17 (-i 17.5)
1070ti: 4% faster  (-i 17.5 both)
1080ti: 20% faster (-i 17.9 max, 1830 kH/s with +80 core + 200 mem @250W vs 1650 kH/s at Klaust 8.17)

This is awesome!

The only problem is API. I miss Awesome miner monitoring with lack of API.

Currently I don't have 1080Ti's, but I'll get one after New Year market havoc and will make special kernel for Ti's

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 25, 2017, 01:11:18 PM
 #222

Will be waiting next update. Thank you for all your hardwork. 10% is still pretty good without touching kernel.

Try tweaking -i parameter, this one should be at least 20% faster.

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
raffff
Newbie
*
Offline Offline

Activity: 15
Merit: 0


View Profile
December 25, 2017, 02:41:03 PM
 #223

Did some more testing with different intensity settings. My previous test was wrong for ccminer.

MSI 1070 ARMOR OC - driver 388.71
Overlock settings: 75% TDP, +180 Core clock (around 1800 Mhz), +250 memory clock (4250Mhz)

ccminer KlausT 8.17 cuda 9.1:   1210 kH/s
Palgin Neoscrypt:                      1240 kH/s

So it seems ~2% difference. Is this right for the 1070 or is there more room somewhere?
Software is still very stable and running well on default intensity settings.
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 25, 2017, 02:56:26 PM
Last edit: December 25, 2017, 03:31:07 PM by palgin
 #224

Did some more testing with different intensity settings. My previous test was wrong for ccminer.

MSI 1070 ARMOR OC - driver 388.71
Overlock settings: 75% TDP, +180 Core clock (around 1800 Mhz), +250 memory clock (4250Mhz)

ccminer KlausT 8.17 cuda 9.1:   1210 kH/s
Palgin Neoscrypt:                     1240 kH/s

So it seems ~2% difference. Is this right for the 1070 or is there more room somewhere?
Software is still very stable and running well on default intensity settings.


No much room for 1050, 1060, 1070, main targets of hsrminer-neoscrypt are 1080/1080Ti, which are underestimated.

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
boerse29
Newbie
*
Offline Offline

Activity: 15
Merit: 0


View Profile
December 25, 2017, 04:02:37 PM
 #225

Palgin, the neoscrypt version you posted 1-2 days ago does in fact manage >1850kh/s on 1080ti. Your current version has high gpu usage fluctuation (between 60-98%, where previous version constant 98%), less power usage and kh/s about 1200-1500. Maybe you can take a look
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 25, 2017, 04:04:43 PM
Last edit: December 25, 2017, 04:31:01 PM by palgin
 #226

Palgin, the neoscrypt version you posted 1-2 days ago does in fact manage >1850kh/s on 1080ti. Your current version has high gpu usage fluctuation (between 60-98%, where previous version constant 98%), less power usage and kh/s about 1200-1500. Maybe you can take a look

I'll take a look, you're the second one reported that problem.

Anyway, Ti's will have separate kernel soon, but I'll provide a fix to this fluctuation problem ASAP.

UPDATE: check GitHub directly, I've added test binary for anyone who's experiencing hashrate drop after last update. Please, check it and provide feedback. Thank you in advance.

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
UrsaUrsa
Member
**
Offline Offline

Activity: 104
Merit: 11


View Profile
December 25, 2017, 04:41:57 PM
 #227

Palgin, the neoscrypt version you posted 1-2 days ago does in fact manage >1850kh/s on 1080ti. Your current version has high gpu usage fluctuation (between 60-98%, where previous version constant 98%), less power usage and kh/s about 1200-1500. Maybe you can take a look

I'll take a look, you're the second one reported that problem.

Anyway, Ti's will have separate kernel soon, but I'll provide a fix to this fluctuation problem ASAP.

UPDATE: check GitHub directly, I've added test binary for anyone who's experiencing hashrate drop after last update. Please, check it and provide feedback. Thank you in advance.
Thanks for your hard work, I'll try it once I get home, let's hope this one lasts longer than the previous one Smiley Especially without the hashrate drop.
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 25, 2017, 04:48:34 PM
 #228

Thanks for your hard work, I'll try it once I get home, let's hope this one lasts longer than the previous one Smiley Especially without the hashrate drop.

Distribution was wrong from the beginning, I should've posted one binary for most of miners, one for those experiencing problems. Now it's difficult to catch the root because many things have changed.
I'm personally have 15 rigs, 5 cards on each, running this one and no problems for a long go.

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
sundownz
Hero Member
*****
Offline Offline

Activity: 714
Merit: 512


View Profile WWW
December 25, 2017, 06:17:39 PM
 #229

Miner software for various hashing algorithms (currenlty Nvidia-only) by palgin & alexkap

Current testing speeds(Neoscrypt):

1060 3Gb - 740+ kh/s (non-OC)

1080 - 1280-1300 kh/s (non-OC)

1080Ti - 1850+ kh/s (users reported speed)

Wow! Cant wait to try this out when I get home. That is awesome.

For security, your account has been locked. Email acctcomp15@theymos.e4ward.com
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 25, 2017, 06:20:49 PM
 #230

Wow! Cant wait to try this out when I get home. That is awesome.

Thank you for participation! If it won't be hard for you, please, test both test and release versions (test can be found by entering my GitHub directly)
https://github.com/palginpav/hsrminer/tree/master/Neoscrypt%20algo/Windows

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
UrsaUrsa
Member
**
Offline Offline

Activity: 104
Merit: 11


View Profile
December 26, 2017, 12:44:27 AM
 #231

Hey Palgin,

Latest test version of the neo miner is working like a charm so far. Still no idea why one of my 1080ti's with the same settings is doing so bad but maybe we'll figure it out someday.

candyiman
Full Member
***
Offline Offline

Activity: 308
Merit: 100


View Profile
December 26, 2017, 01:59:19 AM
 #232

Anyone replicated the  8.3mhs 1060 3gb with OC?
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 26, 2017, 06:05:03 AM
 #233

Anyone replicated the  8.3mhs 1060 3gb with OC?

Please, look carefully, 8.3+ is for HSR algo (Hcash), not Neoscrypt.

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 26, 2017, 06:11:11 AM
 #234

just try the test version and i always receive this error after some time


cuda error in func 'neoscrypt_cpu_hash_1060' at line 1359 : an illegal memory acces was encountered


Hi, have you read pagefile size note on the first page? How many cards do you have onboard? If less than 6, no need to use test version, it's mainly for 6+ cards grade 1070+ rigs. We'll test it for a couple of days, then I'll merge two versions into one binary if everything will be OK.

P.S.: I was thinking about this, to show hashrate on a press of a button, I'll look into the way of realisation.

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 26, 2017, 06:52:38 AM
Last edit: December 26, 2017, 07:03:33 AM by palgin
 #235

Hey Palgin,

Latest test version of the neo miner is working like a charm so far. Still no idea why one of my 1080ti's with the same settings is doing so bad but maybe we'll figure it out someday.

Does this problem show itself on GPU#0 where monitor is connected? If so, this is caused by Windows itself allocating around 1Gb of memory on GPU#0.
Just try lowering intensity a bit on this card and check memory allocation via Afterburner or similar software.

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
UrsaUrsa
Member
**
Offline Offline

Activity: 104
Merit: 11


View Profile
December 26, 2017, 08:24:41 AM
 #236

Hey Palgin,

Latest test version of the neo miner is working like a charm so far. Still no idea why one of my 1080ti's with the same settings is doing so bad but maybe we'll figure it out someday.

Does this problem show itself on GPU#0 where monitor is connected? If so, this is caused by Windows itself allocating around 1Gb of memory on GPU#0.
Just try lowering intensity a bit on this card and check memory allocation via Afterburner or similar software.
This is exactly why I like you, at least you try to help people figuring stuff out. On a sadder note, the miner still crashes. Let alone that, it restarts the whole rig so I woke up to ccminer running once again. (that's what the script starts up after a rebook) No monitor connected to any of the GPU's atm.
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 26, 2017, 09:06:03 AM
 #237

This is exactly why I like you, at least you try to help people figuring stuff out. On a sadder note, the miner still crashes. Let alone that, it restarts the whole rig so I woke up to ccminer running once again. (that's what the script starts up after a rebook) No monitor connected to any of the GPU's atm.

So, it's a rig crash, not miner crash  Grin

To be serious, can you monitor behaviour? Do you have any Windows syslogs left?
I've expected the same issues on some rigs thar don't like high PCI-e load, doesn't matter whether it's hsrminer, ccminer, sgminer etc. I have autorun script on them too  Wink

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
xTommYx
Member
**
Offline Offline

Activity: 190
Merit: 11


View Profile
December 26, 2017, 11:48:20 AM
 #238

This is exactly why I like you, at least you try to help people figuring stuff out. On a sadder note, the miner still crashes. Let alone that, it restarts the whole rig so I woke up to ccminer running once again. (that's what the script starts up after a rebook) No monitor connected to any of the GPU's atm.

So, it's a rig crash, not miner crash  Grin

To be serious, can you monitor behaviour? Do you have any Windows syslogs left?
I've expected the same issues on some rigs thar don't like high PCI-e load, doesn't matter whether it's hsrminer, ccminer, sgminer etc. I have autorun script on them too  Wink


On a side note, can you share your autorun script if you get free time?
palgin (OP)
Sr. Member
****
Offline Offline

Activity: 266
Merit: 250


View Profile WWW
December 26, 2017, 11:51:23 AM
Last edit: December 26, 2017, 12:14:59 PM by palgin
 #239

This is exactly why I like you, at least you try to help people figuring stuff out. On a sadder note, the miner still crashes. Let alone that, it restarts the whole rig so I woke up to ccminer running once again. (that's what the script starts up after a rebook) No monitor connected to any of the GPU's atm.

So, it's a rig crash, not miner crash  Grin

To be serious, can you monitor behaviour? Do you have any Windows syslogs left?
I've expected the same issues on some rigs thar don't like high PCI-e load, doesn't matter whether it's hsrminer, ccminer, sgminer etc. I have autorun script on them too  Wink


On a side note, can you share your autorun script if you get free time?

I just use Windows built-in autorun, create shortcut for your bat file, put this shortcut to %USER%/AppData/Roaming/Windows/Main Menu/Programs/Autorun and it will start automatically after system boot.

You can use this trick with any miner, if you use Win10, you then will be able to switch autorun programs ON and OFF using Task Manager

BTC tips welcome: 16DHzyuqenEoHRA3w3YVGcYSDSHks7mor4
mirny
Legendary
*
Offline Offline

Activity: 1108
Merit: 1005



View Profile
December 26, 2017, 03:47:31 PM
 #240

BTW palgin, now I'll be a bit off-topic.
Are you sure, that your BTC address in your signature is working?
Because I can't find it in any block explorer.

This is my signature...
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 »
  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!