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

[eXtremal]

I think the NVIDIA private farmers already got a 120Sol/s miner / card for the zcash harvesting.

But with 48sols/s on GTX1070 this miner became fastest public.

[1713963592]

1713963592

[1713963592]

1713963592

[1713963592]

1713963592

[1713963592]

1713963592

[1713963592]

1713963592

[ioglnx]

Sadly for me the Genoil modification lead to a crash on Windows 10 as well as 7 on nvidia and amd rigs.
Stick with the NHEQMiner for the moment since i get there around 300h/s with my nvidia its okay to run 2 instances

[cryptomined]

Would be great if someone makes 1 complete update with recent advancements

i second that and hope an update with all the recent optimizations becomes available soon, thanks
also hoping something can be done when a GPU gets dropped, some sort of auto-restart GPU function

But with 48sols/s on GTX1070 this miner became fastest public.

Hi Extremal, i think i ended up sending a lot of miners your way on launch day  
russians were bombing me saying you weren't paying up but i told them not to rush to conclusions and silenced them

Sadly for me the Genoil modification lead to a crash on Windows 10 as well as 7 on nvidia and amd rigs.
Stick with the NHEQMiner for the moment since i get there around 300h/s with my nvidia its okay to run 2 instances

you get 50 sols/s each card with NHEQminer?

i wasnt able to overclock more than 1 card as it wont give me settings in nvidia-settings for the 1x riser cards.. and it wouldnt even let me use my overclock settings when I ran the miner anyway.. any idea what I can do to fix that?

also i get this error in majaro trying to run silentarmy with my 780m's
Solver 0.0: unexpected banner "/home/wolf/Downloads/silentarmy-master/sa-solver: /opt/cuda/lib64/libOpenCL.so.1: no version information available

[avirdoz]

+1 windows support

[nwfella]

Have been running Silentarmy v4 steady now for past 5.75 hours without so much as a hiccup with comparable hashrate (~48sol/sec on 2x 4GB cards, 270x & 280x)

Keep up the great work fellas.

[ioglnx]

Have been running Silentarmy v4 steady now for past 5.75 hours without so much as a hiccup with comparable hashrate (~48sol/sec on 2x 4GB cards, 270x & 280x)

Keep up the great work fellas.

windows or linux?

[eXtremal]

RX480 with amdgpu-pro 16.30

Total 55.8 sol/s [dev0 54.0] 18 shares
Total 55.3 sol/s [dev0 52.4] 18 shares
Total 55.6 sol/s [dev0 54.7] 18 shares
Total 55.9 sol/s [dev0 55.7] 18 shares
Total 55.0 sol/s [dev0 55.7] 18 shares
Total 55.5 sol/s [dev0 56.2] 18 shares
Total 55.2 sol/s [dev0 56.1] 19 shares
Total 54.6 sol/s [dev0 54.8] 19 shares
Total 54.9 sol/s [dev0 55.3] 19 shares
Total 55.1 sol/s [dev0 53.1] 19 shares
Total 54.4 sol/s [dev0 52.6] 19 shares

Kernel: http://coinsforall.io/distr/input.cl.coll1 NVidia also have speedup.
I reduced number of collisions to found from 5 to 1, it seems 5 is too much, need mrb's comments.

[kenshirothefist]

mrb, you still haven't pulled these two?

https://github.com/mbevand/silentarmy/pull/17
https://github.com/mbevand/silentarmy/pull/39

There are many miners, trying to use your miner on NiceHash, but are having issues because of these two issues (and on some others pools as well, because of missing client.reconnect method).

Anyway, if anyone of you would like to mine on NiceHash with silentarmy you can build from our fork https://github.com/nicehash/silentarmy where these two issues are already solved.

Thanks!

[nerdralph]

RX480 with amdgpu-pro 16.30

Total 55.8 sol/s [dev0 54.0] 18 shares
Total 55.3 sol/s [dev0 52.4] 18 shares
Total 55.6 sol/s [dev0 54.7] 18 shares
Total 55.9 sol/s [dev0 55.7] 18 shares
Total 55.0 sol/s [dev0 55.7] 18 shares
Total 55.5 sol/s [dev0 56.2] 18 shares
Total 55.2 sol/s [dev0 56.1] 19 shares
Total 54.6 sol/s [dev0 54.8] 19 shares
Total 54.9 sol/s [dev0 55.3] 19 shares
Total 55.1 sol/s [dev0 53.1] 19 shares
Total 54.4 sol/s [dev0 52.6] 19 shares

Kernel: http://coinsforall.io/distr/input.cl.coll1 NVidia also have speedup.
I reduced number of collisions to found from 5 to 1, it seems 5 is too much, need mrb's comments.

I've just finished testing this, and simply changing the collisions array from 5 to 1 gives a few percent increase in performance.

This also shows that there is the potential for additional performance increases by improving the pruning of duplicate indices.  For a couple days I've been trying to think of a way to do this quickly (i.e. without expanding indexes), but don't have a solution yet.

Also tested reducing the collision search space for later rounds using the formula:

Code:

min(cnt, (uint)(NR_SLOTS -(uint)(round-4)))

No reduction in the number of solutions generated, but no material improvement in speed either.

[laik2]

I see solver has been tested on Xeon Phi.
Any performance feedback?

[laik2]

No idea if it will work (or isn't already implemented)

Quote from: pallas on November 06, 2016, 04:55:27 PM
6 iterations per second CAN lead to 140 sol/s.
Just run multiple nonces per iteration.
It all depends on how you define an "iteration".

PS:
Would be great if someone makes 1 complete update with recent advancements

Am I right to assume that currently the best platform look to be Ubuntu 16 + AMD 16.40 drivers? Just because 16.40 don't work on Ubuntu 14?

Had anyone tried the usage of opencl files from 16.40 on Ub14?
Or, the more tricky way - get compiled binaries from Ub16+16.40 and run on Ub14?

I receive better results with kernel 4.8(OSS amdgpu driver) and amdgpu-pro packages without kernel driver itself)
Sapphire Nitro RX480 8GB ElioVP mod x2 - 123S/s peak, average 110/112.

[isamamin]

I see some peoples get about 165 sols/s  +-  using genoil miner with driver 16.10.3 
but the card should be [8GB] ( he test with RX 470 ).
kernel : zec-sa.cl
comand :

Code:

genoil.exe -k zec zec zec zec -c zec.suprnova.cc:2142 -u user.x1222 -p xnxx -i 20 -w 64 -P 0 -f 2 -z 0

give it try!
 

[isamamin]

I see some peoples get about 165 sols/s  +-  using genoil miner with driver 16.10.3  
but the card should be [8GB] ( he test with RX 470 ).
kernel : zec-sa.cl
comand :

Code:

genoil.exe -k zec zec zec zec -c zec.suprnova.cc:2142 -u user.x1222 -p xnxx -i 20 -w 64 -P 0 -f 2 -z 0

give it try!
 

it  works great but no updates  that's the problem it has major crashes etc but it does work and good if you can deal with the bugs it has .

for CRASH
.bat file

Code:

:loop
genoil.exe -k zec zec zec zec -c zec.suprnova.cc:2142 -u user.x1222 -p xnxx -i 20 -w 64 -P 0 -f 2 -z 0
goto loop

[laik2]

Here I have put a quick howto for linux newbies to squeeze the max of their systems.

Please README.txt twice!

https://drive.google.com/drive/folders/0B72yKpOokCMcVnV5LWNMS2ltYmM?usp=sharing

Please feel free to contact me if you have any problems following or executing this howto.

Stay tuned for new optimizations.

[Genoil]

I have a 0.6.2 up that should be more stable, but it hasn't been tested for very long.

I have been doing the Windows port of SA4 yesterday, maybe now is a good time to get back to my own miner.

[toptek]

I have a 0.6.2 up that should be more stable, but it hasn't been tested for very long.

I have been doing the Windows port of SA4 yesterday, maybe now is a good time to get back to my own miner.

ty i mean that you don't know how much.

[rednoW]

After latest Claymore results it is a hard times for open source devs ))

[laik2]

Here I have put a quick howto for linux newbies to squeeze the max of their systems.

Please README.txt twice!

https://drive.google.com/drive/folders/0B72yKpOokCMcVnV5LWNMS2ltYmM?usp=sharing

Please feel free to contact me if you have any problems following or executing this howto.

Stay tuned for new optimizations.

nice to bad AMD hasn't release driver support for the rest of there cards that are supported in the windows diver so we can use any card with ubuntu 16 which to me is the best OS to use if you go Linux .

I can confirm that RX 480 , R9 lines and R7 APU's work with this driver without any issues 48 hours straight.

[c2n14]

I see some peoples get about 165 sols/s  +-  using genoil miner with driver 16.10.3 
but the card should be [8GB] ( he test with RX 470 ).
kernel : zec-sa.cl
comand :

Code:

genoil.exe -k zec zec zec zec -c zec.suprnova.cc:2142 -u user.x1222 -p xnxx -i 20 -w 64 -P 0 -f 2 -z 0

give it try!
 

Where to find this updated kernel?

[adaseb]

I see some peoples get about 165 sols/s  +-  using genoil miner with driver 16.10.3 
but the card should be [8GB] ( he test with RX 470 ).
kernel : zec-sa.cl
comand :

Code:

genoil.exe -k zec zec zec zec -c zec.suprnova.cc:2142 -u user.x1222 -p xnxx -i 20 -w 64 -P 0 -f 2 -z 0

give it try!
 

Where to find this updated kernel?

I get 30Sol/s but i got 4GB