SILENTARMY v5: Zcash miner, 115 sol/s on R9 Nano, 70 sol/s on GTX 1070

[toptek]

maybe there should be an official silentarmy thread? v5 says they were gonna go for windows support but between the bullshit of claymore jackin for beats and im sure getting bugs and improvements + windows support it will take a minute. just trollin

down load one of these
https://mega.nz/#F!iwYQhI6S!2vl0D3jz3fH3-bZxHUurBg

the latest version is  SILENTARMY-v5-Cygwin64-r6

 down load this

https://cygwin.com/setup-x86_64.exe

then unzip

SILENTARMY-v5-Cygwin64-r6  then read WINDOWS.txt

 it tells you what package you need form cygwin

your set .

[1714623150]

1714623150

[1714623150]

1714623150

[1714623150]

1714623150

[1714623150]

1714623150

[1714623150]

1714623150

[toptek]

Silentarmy vs Optiminer results

I'm running ubuntu with 6x rx 470 8GB

With SA I averaged 520 sols/s using 700W
With Optiminer I'm averaging 660 sols/s using 740W

That is about a 27% increase in speed and this is all with the 15% dev fee included already.

I'm going to have to stick with Optiminer until SA releases a faster Linux miner.

Wow, 15% devfee is insane.  And so many whined about Claymore devfee, even when miner was 100% faster.  I was assuming the devfee on that miner was crazy high since he wouldn't tell you what it was.  Some report Optiminer as unstable, and it is Linux only and harder to set up. Are you comparing solves over time at pool?

Silentarmy vs Optiminer results

I'm running ubuntu with 6x rx 470 8GB

With SA I averaged 520 sols/s using 700W
With Optiminer I'm averaging 660 sols/s using 740W

That is about a 27% increase in speed and this is all with the 15% dev fee included already.

I'm going to have to stick with Optiminer until SA releases a faster Linux miner.

Wow, 15% devfee is insane.  And so many whined about Claymore devfee, even when miner was 100% faster.  I was assuming the devfee on that miner was crazy high since he wouldn't tell you what it was.  Some report Optiminer as unstable, and it is Linux only and harder to set up. Are you comparing solves over time at pool?

The dev fee is a bit ridiculous but still not bad enough to not use his miner. I have been running the miner for about 4hrs now with with no issues. And I am using flypool to compare the two.

If it works, then go ahead and use it, if you get more profit after devfee, its better for you.  I would try it if it were on Windows.  I am just observing no one has complained about the 15% devfee, yet everyone complained about 2.5% when there were a ton of slow, unstable, and/or pool locked miners at the time.

That's why i haven't tired it . and I'm not going to . I wasn't complaining about claymore  fee so much as i was give us a choice or way around it. he did !!! . imo  to late. and his miner at the time was the only working windows miners worth using and after he put in the off flag for the fee i left it on.  ).

[boestin]

Toptek you are my hero

[toptek]

i don't how to take that JK   
 
thanks top.

[bardacuda]

I think in terms of speed claymore and sav5 are almost the same, but with share finding, I can conclude that claymore does find more. I tracked all the shares with SA5 and compare it with the logs on claymore miner and the difference is 27% for 24 hours.

I'm seeing the opposite.

[giagge]

maybe there should be an official silentarmy thread? v5 says they were gonna go for windows support but between the bullshit of claymore jackin for beats and im sure getting bugs and improvements + windows support it will take a minute. just trollin

down load one of these
https://mega.nz/#F!iwYQhI6S!2vl0D3jz3fH3-bZxHUurBg

the latest version is  SILENTARMY-v5-Cygwin64-r6

 down load this

https://cygwin.com/setup-x86_64.exe

then unzip

SILENTARMY-v5-Cygwin64-r6  then read WINDOWS.txt

 it tells you what package you need form cygwin

your set .

No have improvement on gtx 1070 with r6 , i have same result with r5 .

[svetliopi44]

how to draw at Nice hash
bat a file
 directly to mine bitkoin
 

[laik2]

I think in terms of speed claymore and sav5 are almost the same, but with share finding, I can conclude that claymore does find more. I tracked all the shares with SA5 and compare it with the logs on claymore miner and the difference is 27% for 24 hours.

I'm seeing the opposite.

Sharerate looks the same to me...Needs more verbose output about shares(submit,invalid,rejected) in silentarmy to conclude something.

[nerdralph]

SA5 has no more easy optimizations left; the current version maxes out the memory bandwidth of the card.

Does it also max out the PCI-E bus bandwidth? I'm trying to understand why my Nitro+ RX470 (at 75~80 Sol/s) is finding as many shares as my R9 295x2 (2 x 90 Sol/s)...  Any idea?

With the on-GPU solution pruning it barely uses any PCI-e bus bandwidth.  My comment about maxing out the memory bandwidth is a simplification of a complex problem.  The memory path includes L1 & L2 cache as well as the external GDDR5.  I believe parts of the current performance bottleneck are due to L1/L2 cache thrashing.  There's at least a couple ways of solving the problem, but none (that I can think of) are easy.

[marvykkio]

I installed python 3, but when I open the terminal what is the command to be run?
 

[laik2]

I installed python 3, but when I open the terminal what is the command to be run?
 

Edit start.bat with your credentials(pytho3 for windows is not dependency - read carefully - install cygwin64 in path c:\cygwin64)

[laik2]

SA5 has no more easy optimizations left; the current version maxes out the memory bandwidth of the card.

Does it also max out the PCI-E bus bandwidth? I'm trying to understand why my Nitro+ RX470 (at 75~80 Sol/s) is finding as many shares as my R9 295x2 (2 x 90 Sol/s)...  Any idea?

With the on-GPU solution pruning it barely uses any PCI-e bus bandwidth.  My comment about maxing out the memory bandwidth is a simplification of a complex problem.  The memory path includes L1 & L2 cache as well as the external GDDR5.  I believe parts of the current performance bottleneck are due to L1/L2 cache thrashing.  There's at least a couple ways of solving the problem, but none (that I can think of) are easy.

We all have faith in you

[jk_14]

Hi, mrb

Could you implement some very usefull, ccminer-like option: -S --syslog

  -S, --syslog          use system log for output messages

that would be nice for KopiemTu support, i.e. for monitoring

It's low priority, but I will consider it. In the mean time I am happy to accept patches from contributors.

Ok, that is my patch, then

Code:

root@kopiemtu:~# diff /opt/miners/silentarmy/silentarmy silentarmy.orig
13d12
< import syslog
266,271c265
<
<             if self.opts.syslog:
<                 syslog.syslog("Total %s sol/s [%s] %d share%s" % \
<                     (rate_avg, info_gpus, shares, '' if shares == 1 else 's'))
<             else:
<                 print("Total %s sol/s [%s] %d share%s" % \
---
>             print("Total %s sol/s [%s] %d share%s" % \
273c267
<                 sys.stdout.flush()
---
>             sys.stdout.flush()
401,402d394
<             if self.opts.syslog:
<                print('\nNow logging to syslog...')
404,405d395
<
<
566,569d555
<     parser.add_option(
<             "--syslog",
<             dest="syslog", action="store_true",
<             help="use system log for output messages")
root@kopiemtu:~#

whole modified file:
http://www.mbl.cash/silentarmy

Please apply it, thanks!

[ivomm]

I installed cygwin and followed the instructions for windows. It worked for my 280x cards but not for 390 cards. Any ideas?

[Kubuxu]

SA5 has no more easy optimizations left; the current version maxes out the memory bandwidth of the card.

Does it also max out the PCI-E bus bandwidth? I'm trying to understand why my Nitro+ RX470 (at 75~80 Sol/s) is finding as many shares as my R9 295x2 (2 x 90 Sol/s)...  Any idea?

With the on-GPU solution pruning it barely uses any PCI-e bus bandwidth.  My comment about maxing out the memory bandwidth is a simplification of a complex problem.  The memory path includes L1 & L2 cache as well as the external GDDR5.  I believe parts of the current performance bottleneck are due to L1/L2 cache thrashing.  There's at least a couple ways of solving the problem, but none (that I can think of) are easy.

I was profiling it yesterday, the limiting factor right now is LDS. To optimize it, total memory accessed by the round kernel wave would have to be reduced. Currently it falls into about 16K. CodeXL says that utilization during rounds is 12.5% due to this. I can send you more data if you want or you can profile it yourself with CodeXL 2.2.

[laik2]

I installed cygwin and followed the instructions for windows. It worked for my 280x cards but not for 390 cards. Any ideas?

It's probably due to driver issues...280x uses different driver than 390...please use GPU-Z to find if there are driver issues.

[gollygosh]

Tried using the windows port by zawawa and this was the result
upload an image

I'm using windows 7 64bit ultimate and installed cygwin64 on to main C:\ drive with Python3: Py3K language interpreter selected.
Error was with default start.bat being run. I have 2 cards in the rig 1 RX470 and the other is an RX 480 with latest AMD drivers.

Anyone know what I did wrong?

Thanks for any help

[ivomm]

I installed cygwin and followed the instructions for windows. It worked for my 280x cards but not for 390 cards. Any ideas?

It's probably due to driver issues...280x uses different driver than 390...please use GPU-Z to find if there are driver issues.

I have no driver issues. The cards work perfect with claymore miner.

P.S. I got the same message after waiting 6-7 min as the above post.

[richx]

can anyone help. mining with 1 1070 card i get ~70/s, but with 6x1070 only ~150?
my cmd line:
py silentarmy --use 0 -c stratum+tcp://cn1-zcash.flypool.org:3333 -u t1KmpFfV7iZyNefokBgbMUpSYyMZX2qm7cr.A1 --instances=1

6 cards
py silentarmy --use 0,1,2,3,4,5 -c stratum+tcp://cn1-zcash.flypool.org:3333 -u t1KmpFfV7iZyNefokBgbMUpSYyMZX2qm7cr.A1 --instances=1

[Maxikosw]

SILENTARMY-v5-Cygwin64-r6.zip i get 14 SOL. R9 Nano...