Topic: [ANN] lolMiner-mnx, Mining MinexCoin(MNX) New version: v0.4 (July 22) - page 23. (Read 30772 times)

ver. 0.3 is really great.
It has a higher hash rate and has not crashed more than 20 hours.

Is this true?

On my ubuntu 17.10 machine with 32GB physical memory I just have a 16GB swap partition and have a GTX1060 3GB plus on-board Quadro M1200 4GB so going by your reckoning I should set swap size to be 40GB ?

I'll try adding another swap file sized 24GB and see if there is any difference in performance.

As mentioned in telegram yesterday, with v0.3 on my Ubuntu 17.10 gen7 i7 machine the "--list-devices" command produces no output, however I can run it using the config used with v0.22 but the Sol/s are much lower for intel CPU and the laptops onboard Nvidia Quadro M1200 but the (not overclocked) GTX1060 3GB connected via TB3 in an eGPU box is only marginally faster with v0.3

EG:

$ ./lolMiner-mnx_v03 --list-devices
-------------------------------------------------
          Welcome to lolMiner-mnx v0.3

   This software is only meant for pool mining
    with the Equihash 96/5 algorithm (MARS)

For usage instructions add -h to the commandline.
   For support visit our bitcointalk thread.

          Made by Lolliedieb, March 2018
-------------------------------------------------
All available OpenCL platforms and devices:


$ ./lolMiner-mnx_v022 --list-devices
-------------------------------------------------
          Welcome to lolMiner-MNX v0.22

   This software is only meant for pool mining
    with the Equihash 96/5 algorithm (MARS)

For usage instructions add -h to the commandline.
   For support visit our bitcointalk thread.

        Made by Lolliedieb, Feburary 2018
-------------------------------------------------
All available OpenCL platforms and devices:

Platform 0: NVIDIA CUDA
    Device 0: GeForce GTX 1060 3GB (GPU)
    Device 1: Quadro M1200 (GPU)
Platform 1: Intel(R) OpenCL
    Device 0: Intel(R) HD Graphics (GPU)
    Device 1: Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz (CPU)


Average speeds of each device (excluding the Intel HD graphics as this causes all CPU cores to run near 100%) and on each version using mostly a default lolMiner config are:


NVIDIA GeForce GTX 1060 3GB (not over clocked): ~10.1KSol/s @ 96W with v0.22 and ~10.3KSol/s @ 100W with v0.3
NVIDIA Quadro M1200: ~3.6KSol/s with v0.22 and ~2.1KSol/s with v0.3
Intel i7-7820HQ CPU: ~600Sol/s with v0.22 and max rate about ~581Sol/S with v0.3 (no core affinity set, top shows it using only 1% CPU)

I am not sure why the v0.3 "--list-devices" command shows nothing but the openCL libraries I am running have been compiled using an Intel supplied script available here:
https://software.intel.com/en-us/articles/sdk-for-opencl-gsg

But using a slightly modified version of the "experimental" script was used (https://software.intel.com/file/593325/download) to make it work with Ubuntu 17.10 (kernel 4.13.0-36-generic)

The Intel page also mentions a python System Analyzer Utility available here:
https://software.intel.com/en-us/articles/mss-sys-analyzer-linux

And the output from this on my machine (that does not have Media Server Studio Installed) is:

# sys_analyzer_linux.py
--------------------------
Hardware readiness checks:
--------------------------
 [ OK ] Processor name: Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz
--------------------------
OS readiness checks:
--------------------------
 [ OK ] GPU visible to OS
--------------------------
Media Server Studio Install:
--------------------------
 [ OK ] user in video group
 [ ERROR ] libva.so.1 not found.  Check LD_LIBRARY_PATH contains '/usr/lib64;/usr/local/lib'
 [ ERROR ] libva not loading Intel iHD
 [ ERROR ] vainfo not reporting codec entry points
 [ ERROR ] could not open /dev/dri/renderD128
 [ OK ] /dev/dri/renderD129 connects to Intel i915
--------------------------
Component Smoke Tests:
--------------------------
 [ ERROR ] no Media SDK include files.  Are Intel components installed?
 [ OK ] OpenCL check:platform:NVIDIA CUDA GPU OK CPU FAIL
platform:Intel(R) OpenCL GPU OK CPU OK

[Device 0] Average speed (5s): 15885.2 sol/s 1070
[Device 1] Average speed (5s): 15335.9 sol/s 1070
[Device 2] Average speed (5s): 10088.4 sol/s 1060 3 gb
[Device 3] Average speed (5s): 15520.5 sol/s 1070
[Device 4] Average speed (5s): 6076.18 sol/s 1050 ti


all works well for 2 days

I have heard from some people that the lower end Nvidia cards GT 1030, GTX 1050 (Ti) and older cards ( 8X00M, GTX 750 Ti and lower)  are slower with the new version and that for GTX 1060 its a close calculation because it is quicker, but consumes more energy as well.

Therefore I will try to build 0.31 soon with a switch to turn back the optimizations if desired.

Thanks all for feedback so far!

Can you please share your config or params. I have p106 6Gb and getting 9200 sol, cannot get more.

Thanks

Getting 18 KSol/s with 1070Ti and 25-28 KSol/s with 1080Ti. Using Ubuntu 16.04, I've experienced a few segfaults but I'm running the miner using systemd with autorestart so this haven't crashed or locked up the computer. Applied same overclock settings as with DSTM miner.

v0.3 is working great for me. The only thing is that for the set-work-batch parameter, I could not use AUTO or HIGH. The miner would automatically crash on these settings. I had to use actual numbers to get it working.

tnx man  nice update :d my 1060 rate up 1000sol    last ver i have 10600sol
but last night i run it and morning speed  0sol !!!  i think because of OC    i set lower OC for cards see what  happend tankx aging ! nice update.  



[Device 0] Average speed (5s): 11346.7 sol/s   1060 6g
[Device 1] Average speed (5s): 11410.1 sol/s    1060 6g
[Device 2] Average speed (5s): 6426.71 sol/s    1050 ti
[Device 3] Average speed (5s): 6386.32 sol/s     1050ti

What am I doing wrong?

@Lolliedieb
I'm wondering if it would be relatively easy to add other Equihash variants to the miner, specifically 192,7 for Zero coin
Right now we have just one miner option, OptiminerZero (and OptiminerEquihash, which is linux only)

Could you look into it?

On ubuntu on GTX1050/GTX1050ti/GTX1060 has increase of 200-400 sol/s. This version more profit for windows.

V0.3 is running for the last 10+ hours on my 2 x 1070s with no crashes and average combined of 34000 sol/s.
Good work!

Lolliedieb,
Thank you for your hard work!

The new version is working much better for me, at least in that it does not crash.

I did not get noticeable speed increase but I am happy with what I am getting.

This is on two GTX1050TI cards (single fan model that powers only from slot)

Average speed (60s): 6716.49 sol/s | 6750.09 sol/s Total: 13466.6 sol/s
---------------------------------------------
[Device 0] Average speed (5s): 6605.88 sol/s
[Device 1] Average speed (5s): 6722.86 sol/s
[Device 0] Average speed (5s): 6714.86 sol/s
[Device 1] Average speed (5s): 6781.24 sol/s
[Device 0] Average speed (5s): 6535.04 sol/s
[Device 1] Average speed (5s): 6699.34 sol/s
[Device 0] Average speed (5s): 6722.17 sol/s
[Device 1] Average speed (5s): 6734.16 sol/s

I did have to dial back my clock speed a bit because I got the 0sol a few times but after reducing the clock it was fine.

I am using MSI Afterburner for clocking and have set core to +75 and Mem +700,, running very stable for many hours now.

MSI displays 1784Mhz GPU clock and 4203Mhz Mem clock

In the config I have enabled:
--set-work-batch HIGH
--set-sleep-duration 0

Thanks for the new version, it really is much faster. I like the option of not detecting work. Can someone pass me a Script or Bat of supervision, so that when the miner closes for that new option, I will re-launch it ....


Also say that the only weird thing that I have detected for now, that when you start it, it does not always do it, many times it closes at the very beginning, and I have to launch it a second time.

That's why I want to control when a card fails with the new option but I do not know of a small script that relaunches the miner's .bat if it detects that it stops or something that can serve

PC does not go to sleep when lolminer is running. It is troublesome to turn off the monitor every time.
Is there an option to turn off this feature?

With v 0.22 Geforce has halved the hash rate. Improved when opening the browser displaying youtube and displaying Remote Desktop.
In 0.3 it was fixed. Closing Remote Desktop does not reduce the hash rate.

Pretty nice update

Unstable work, stock rx580/8gb - every ~5 minutes miner restarts. version 03

25000 sol for 1080 ti

i love you