Bitcoin Forum
October 31, 2024, 06:30:52 PM *
News: Bitcoin Pumpkin Carving Contest
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: Big microCoin settings thread  (Read 2247 times)
cryptofella (OP)
Newbie
*
Offline Offline

Activity: 24
Merit: 0


View Profile
April 22, 2014, 12:51:05 PM
Last edit: May 02, 2014, 06:41:31 PM by cryptofella
 #1

Hi dear microCoiners,

MicroCoin is starting to get big and I think it's good to share our miner settings to make sure everyone is getting all the possible power from their rigs.

Current diff is about 0.2. You can calculate how many hours it takes to solve a block:

Code:
Difficulty*2^32/khashes/3.6*10^-6

I'm using this miner for GPU:
https://github.com/Thirtybird/YACMiner/releases/download/3.5.0-yac2/yacminer-3.5.0-yac2-x86.zip

My .bat file for R9 290X is following:
Code:
yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 3532 -R 10496 -o 127.0.0.1:3333 -u user -p password --gpu-engine 1000 --gpu-memclock 1500 --scan-time 2 --expiry 4

I'm getting 50 khash per card with this.

Also a CPU miner can be found from there:
32: https://github.com/Thirtybird/cpuminer/releases/download/v2.4.0/minerd-yacoin_stratum_x86_2_4_0.zip
64: https://github.com/Thirtybird/cpuminer/releases/download/v2.4.0/minerd-yacoin_stratum_x64_2_4_0.zip

Feel free to post your own configs as well.

~cf~
pandas_talks
Full Member
***
Offline Offline

Activity: 134
Merit: 100


View Profile
April 22, 2014, 08:43:39 PM
 #2

this one is also useful
Hello Guys,

I can share some mining configs with you.

First you have to download Yacminer - https://github.com/Thirtybird/YACMiner/releases/download/3.5.0-yac2/yacminer-3.5.0-yac2-x86.zip
These configs works on systems with more than 4GB RAM.

Config for 290x - yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 3532 -R 10496
Config for 290 - yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 3420 -X 32
Config for 280x - yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 2590 -R 7680
Config for 280 - yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 2394 -X 32
Config for 270x - yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 1648 -R 4864
Config for 7950 - yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 2394 -X 32
Config for 7970 - yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 2590 -R 7680

Tips on tuning your GPU to mine Microcoin
--buffer-size XXXX (-B XXXX):
This setting sets the size of the OpenCL buffer to allocate to each thread. This option overrides --thread-concurrency, and should be used instead of it if your miner supports it. This value should be based on the amount of available memory on your GPU. This setting is NOT affected by lookup-gap meaning that other tuning parameters that you want to test will not force you to recalculate your buffer-size. The only setting that will make you want to change this is the --gpu-threads (-g) setting. As buffer-size is the amount of memory to allocate per thread, if you increase the number of threads you are launching, you will want to decrease the buffer-size of each thread to ensure all of the memory resides in Dedicated GPU memory instead of rolling over to Dynamic.
As the NFactor for Microcoin increases, the amount of memory that is needed to perform a single calculation rises. The more memory that can be allocated per thread, the better the performance that typically can be achieved.
SUMMARY: adjust this value first - Run GPUz of HWInfo to monitor the amount of dedicated memory used on the card until it moves into the dynamic column, then reduce it back down. YACMiner will tell you at the start how much memory has been allocated per thread

--lookup-gap X:
This tunes a compromise between ram usage and performance. Performance peaks at a gap of 2, but increasing the gap may allow you to run at a higher rate as more shaders in the card get utilized. Setting lookup gap overrides the default of 4.
SUMMARY: Start at 2, and try 4 through 8. Adjust your intensity setting (-R/-X/-I) to higher values as the lookup-gap increases.

--worksize XXX (-w 256):
Has a minor effect, should be a multiple of 32 up to 256 maximum. This sets the smallest size of work being sent to the GPU, and on older cards and/or drivers affects the values allowed in Raw Intensity.
SUMMARY: Higher values are better for cards with higher shader counts and 7XXX series cards and newer. Lower values for older cards give finer ranges of Raw Intensity to maximize performance.

--rawintensity XXXX (-R XXXX) This setting allows for a much more preceise intensity setting and can recognize significant gains when mining Microcoin
Note: make sure to use much lower raw-intensity values when you increase cpu threads (-g).
The original Intensity setting is currently used to spawn GPU threads as a simple 2^value setting.

  I:8  =     128 threads
  I:8  =     256 threads
  I:9  =     512 threads
  I:10 =    1024 threads
  I:11 =    2048 threads
  I:12 =    4096 threads
  I:13 =    8192 threads

Notice how the higher settings increase thread count tremendously.
It's now much easier to control thread intensity specifically and it potentially allows for a uniform way of setting the intensity on your system.

Valid values for raw-intensity are safest to be set as a multiple of --worksize. Newer drivers and/or cards may allow for adjustment in smaller increments - in some cases, this can be adjusted in in increments of 1.

--xintensity XX (-X XX)
When you have enough GPU memory allocated to fully utilize all shaders on your GPU, use this for intensity control as it will spawn threads as a multiple of how many shaders are on your GPU. This setting will be best starting at 4, and increases can be had by going up to 32 or as high as 128. Very large values for this may cause instability or driver crashes for very little gain - consider 128 to be the max.
Optional parameters for tuning:
--gpu-threads X (-g X):
Some people who have systems where the GPUs have large amounts of memory and low amounts of system memory will need to run multiple CPU threads so that all of the GPU memory can be allocated. This is because the OpenCL buffer is built in system memory and passed to the GPU. This defaults to 1, however values of 1, 2, 3 & 4 can all produce similar results.
SUMMARY: Run 1 unless your card has more than 2GB of memory and you cannot allocate all of that memory with just 1 thread.
Other available parameters that have better alternatives:
--thread-concurrency XXXXX:
This option may be depreciated in your mining software depreciated. It is still available for those people who like to do math on really high numbers that have to change when any other parameter needs to change. I strongly recommend utilizing the --buffer-size option instead. This setting adjusts the size of the OpenCL buffer to allocate to each thread. This value should be based on the amount of available memory on your GPU. This setting is affected by lookup-gap as well, utilizing the following formula for lookup-gaps of 1,2 & 4

MB per thread = thread-concurrency * 128 / (1024 * lookup-gap)
A good example of this: lookup-gap of 2, thread-concurrency of 8192 = 512 MB

As the NFactor for a Microcoin increases, the amount of memory that is needed to perform a single calculation rises. The more memory that can be allocated per thread, the better the performance that typically can be achieved.
SUMMARY: Use --buffer-size instead

--intensity XX (-I XX)
Just like in bitcoin mining, Microcoin mining takes an intensity, however the scale goes from 0 to 20 to mimic the "Aggression" used in mtrlt's reaper. The reason this is crucial is that too high an intensity can actually be disastrous with Microcoin because it can run out of memory and you will receive HW errors. High intensities start writing over the same ram and it is highly dependent on the GPU, but they can start actually DECREASING your hashrate, or even worse, start producing garbage with HW errors skyrocketing.
SUMMARY: Setting this for reasonable hashrates is mandatory. This setting is not recommended for Microcoin - use --raw-intensity

All the info is from Ultracoin forum.

English ↔ Russian translator
randywald
Hero Member
*****
Offline Offline

Activity: 600
Merit: 500


View Profile
April 24, 2014, 07:25:27 PM
 #3

What kh/s are normal for R9 270, 280X, 290 and 290X? That would be very useful for me because my R9 270 did only about 20-25 kh/s as i tested it the last time. Thank you! :-)

cryptofella (OP)
Newbie
*
Offline Offline

Activity: 24
Merit: 0


View Profile
April 24, 2014, 11:05:38 PM
 #4

What kh/s are normal for R9 270, 280X, 290 and 290X? That would be very useful for me because my R9 270 did only about 20-25 kh/s as i tested it the last time. Thank you! :-)

See comparison with scrypt here:
https://litecoin.info/Mining_hardware_comparison

As you can see 270 gives about half what 290X gives at it's best. My 290X is giving 50 khash with the current n-factor so yours looks good. However, those results could always be improved. You can find suitable memclocks from the link above also and you may want to use AMD Catalyst for tuning. There are a lot of Litecoin related tutorials that could be applied to microCoin as well. DuckDuckGo is our friend here.

Please post your config if you managed to find some improved values.
randywald
Hero Member
*****
Offline Offline

Activity: 600
Merit: 500


View Profile
April 25, 2014, 12:04:32 PM
 #5

Ok, thank you. I always had good settings but i was confused that it dropped from 95 to 20-25 with n-factor increase. I thought with n-factor increase i would get exactly the half hashrate, so i expected about 40-50 for R9 270, but that seems impossible.

cryptofella (OP)
Newbie
*
Offline Offline

Activity: 24
Merit: 0


View Profile
April 25, 2014, 10:37:09 PM
 #6

Ok, thank you. I always had good settings but i was confused that it dropped from 95 to 20-25 with n-factor increase. I thought with n-factor increase i would get exactly the half hashrate, so i expected about 40-50 for R9 270, but that seems impossible.

Yeah, n-factor isn't propotional to hashrate. Whan n-factor increases on May the hashrate on GPU will be even lower and we will see more CPU miners. This is good for network security.
Hilux74
Legendary
*
Offline Offline

Activity: 912
Merit: 1000



View Profile
April 25, 2014, 11:27:04 PM
 #7

My MSI R9 270's for MRC at current Nfactor are doing 37 so you should be able to get better than 25.  What settings are you running?
randywald
Hero Member
*****
Offline Offline

Activity: 600
Merit: 500


View Profile
April 26, 2014, 11:29:20 AM
 #8

I used: yacminer --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 128 -g 1 --lookup-gap 3 --buffer-size 1648 -R 4864

With the n-factor before now i had perfect settings but then i stopped mining after that. The settings were: (--scrypt --Nfmin 4 --Nfmax 30 --StartT 1389028879 --queue 0 --scan-time 2 --expiry 4 --no-submit-stale -w 256 --gpu-engine 1125 --gpu-memclock 1500 -g 1 -I 16 --thread-concurrency 21568)

Can you give me your config? Perhaps i will continue mining if i get good hashrates. Maybe i need to upgrade the RAM on that Rig.

Hilux74
Legendary
*
Offline Offline

Activity: 912
Merit: 1000



View Profile
April 26, 2014, 09:35:54 PM
 #9

yacminer.exe --scrypt-chacha -o stratum+tcp://mrc.blocksolved.com:3310 -u Tongue -p Tongue --nfmin 4 --nfmax 30 --starttime 1389028879 --queue 0 --scan-time 2 --gpu-engine 1100 --gpu-memclock 1200 --rawintensity 4980 --buffer-size 1640 -g 1 -w 256 --lookup-gap 3 --auto-fan --temp-target 75

36.x khash/s @ engine clock 1100
38.x khash/s @ engine clock 1150

Pretty much those settings on all my systems but depending if HW errors show up I have dropped rawintensity down as far as 4640 and even that is still getting ~35khash/s.  Only standout diff I see from your settings is I have w=256 and you have w=128.

Please keep us updated if you are able to increase your hash with these settings.
randywald
Hero Member
*****
Offline Offline

Activity: 600
Merit: 500


View Profile
April 26, 2014, 09:55:14 PM
 #10

Thank you. I will test settings by next week because i've remote access only atm and i'm afraid that the system freezes when i'm doing something wrong ;-)

randywald
Hero Member
*****
Offline Offline

Activity: 600
Merit: 500


View Profile
April 30, 2014, 06:56:34 PM
 #11

Settings didn't work. I always got HW errors, even with -R 4640. I think the R9 270 machine has not enough RAM. I may upgrade it soon.

Now I use my Radeon HD 7950. I can reach ~40 kh/s with -I 12. That's ok for me :-)


Settings:

setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1

microcoinminer.exe -o stratum+tcp://mrc.blocksolved.com:3310 -u x -p x --scrypt --Nfmin 4 --Nfmax 30 --StartT 1389028879 --queue 0 --scan-time 2 --expiry 4 -g 1 -w 256 --gpu-engine 1125 --gpu-memclock 1250 -I 12 --thread-concurrency 35000

randywald
Hero Member
*****
Offline Offline

Activity: 600
Merit: 500


View Profile
April 30, 2014, 10:58:44 PM
 #12

I also tried cpu-mining. Here are the results, if anyone wants to try it:

- Intel Pentium G3220 -> 1 kh/s (2 threads)
- Intel Core i7 980X -> 5.85 kh/s (12 threads)

At N14 and higher i think cpu-mining can be worthwile!

randywald
Hero Member
*****
Offline Offline

Activity: 600
Merit: 500


View Profile
May 01, 2014, 05:02:43 PM
 #13

Is anybody mining with Radeon HD 5870s?

I need correct settings for my Sapphire Radeon HD 5870 Vapor-X. I always find perfect settings for litecoin, vertcoin etc., but for microcoin i'm overwhelmed with all the n-factor changes and HW errors. Cry Would be nice, if someone would post a working config for me.

Thank you!

aikklond
Legendary
*
Offline Offline

Activity: 1148
Merit: 1000


View Profile
May 03, 2014, 06:54:33 PM
 #14

Is it normal to get some HW errors per minute?
Bansheroom
Sr. Member
****
Offline Offline

Activity: 434
Merit: 250


View Profile
May 03, 2014, 07:09:17 PM
 #15

Is it normal to get some HW errors per minute?

HW errors are always bad. Should be zero or at least very close to zero.
Try to lower shaders, thread concurrency, buffers, whatever the miner you use has, slightly in very little steps until the HW errors are gone.

Always get cryptonews, free bitcoins, free altcoins and free mining opportunities: follow @Bansheroom on Twitter
aikklond
Legendary
*
Offline Offline

Activity: 1148
Merit: 1000


View Profile
May 03, 2014, 07:18:02 PM
 #16

Is it normal to get some HW errors per minute?

HW errors are always bad. Should be zero or at least very close to zero.
Try to lower shaders, thread concurrency, buffers, whatever the miner you use has, slightly in very little steps until the HW errors are gone.

 --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 64 -g 1 --lookup-gap 3 --buffer-size 1200 -X 32

these are my settings for all three cards, buffer size has no effect, what should i change?

With what command can I select specific GPU's?
Bansheroom
Sr. Member
****
Offline Offline

Activity: 434
Merit: 250


View Profile
May 03, 2014, 07:27:58 PM
 #17

Is it normal to get some HW errors per minute?

HW errors are always bad. Should be zero or at least very close to zero.
Try to lower shaders, thread concurrency, buffers, whatever the miner you use has, slightly in very little steps until the HW errors are gone.

 --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 64 -g 1 --lookup-gap 3 --buffer-size 1200 -X 32

these are my settings for all three cards, buffer size has no effect, what should i change?

With what command can I select specific GPU's?

I can only answer last part:

--device 0
--device 1

and so on

edit: you should add --rawintensity xxxx and play around with xxxx depending on your gpu.

from a post above:

--rawintensity XXXX (-R XXXX) This setting allows for a much more preceise intensity setting and can recognize significant gains when mining Microcoin
Note: make sure to use much lower raw-intensity values when you increase cpu threads (-g).
The original Intensity setting is currently used to spawn GPU threads as a simple 2^value setting.

  I:8  =     128 threads
  I:8  =     256 threads
  I:9  =     512 threads
  I:10 =    1024 threads
  I:11 =    2048 threads
  I:12 =    4096 threads
  I:13 =    8192 threads

Always get cryptonews, free bitcoins, free altcoins and free mining opportunities: follow @Bansheroom on Twitter
ekrem
Full Member
***
Offline Offline

Activity: 235
Merit: 100



View Profile
May 03, 2014, 07:32:03 PM
 #18

Is it normal to get some HW errors per minute?

HW errors are always bad. Should be zero or at least very close to zero.
Try to lower shaders, thread concurrency, buffers, whatever the miner you use has, slightly in very little steps until the HW errors are gone.

 --scrypt-chacha --nfmin 4 --nfmax 30 --starttime 1389028879 --worksize 64 -g 1 --lookup-gap 3 --buffer-size 1200 -X 32

these are my settings for all three cards, buffer size has no effect, what should i change?

With what command can I select specific GPU's?

please instal  ati driver version 13.12 use yac minner 3.50
randywald
Hero Member
*****
Offline Offline

Activity: 600
Merit: 500


View Profile
July 22, 2014, 03:46:55 PM
Last edit: July 30, 2014, 11:47:36 AM by randywald
 #19

microCoin Mining Instructions

Algorithm: Scrypt-Jane POW mining with modified nFactor
Current nFactor: Time: 1405727099 19 Jul 2014 N: 32768 nFactor: 14




01 - Mining Software


You need a special mining software for microCoin because it doesn't use a standard scrypt algorithm like litecoin or dogecoin.


GPU Mining (AMD):

Several tests have shown that YACMiner 3.5.0-yac2 gives the best hashrate for microCoin.
https://bitcointalk.org/index.php?topic=475569.0

YACMiner - download latest release:
https://github.com/Thirtybird/YACMiner/releases


CPU Mining:

http://microcoin.alcurex.info/CPUminer.zip


02 - Mining Pools


Currently we only have one decent mining pool:


mrc.blocksolved.com


Webpage:
http://mrc.blocksolved.com/

Getting started guide:
http://mrc.blocksolved.com/index.php?page=gettingstarted

Example mining script:
Code:
yacminer -o stratum+tcp://mrc.eu.blocksolved.com:3310 -u websiteUsername.workername -p workerPassword --scrypt --Nfmin 4 --Nfmax 30 --StartT 1389028879 --queue 0 --scan-time 2 --expiry 4


03 - Setup your Miner


Required Hardware:

Operating System: Windows 7 or Windows 8
RAM: 8 GB recommended
GPU: as desired



Settings for your Operating System:

You need to set environment variables to achieve high gpu ram usage - setx commands never worked properly for me, so please do it like i tell you below:

- go to system properties -> advanced -> environment variables (i hope this is correct; i have german windows system)

- like this: http://www.itechtalk.com/thread3595.html

You need to set a new system variable and not user variable

variable name: GPU_MAX_ALLOC_PERCENT
variable value: 100

don't use setx commands in your batch file, this is not needed!



YACMiner Settings for your Hardware:

http://www.ultracoin.net/configgen_raw.html



CPUMiner Settings for your Hardware:

See README.txt in CPUMiner folder to find correct settings for your CPU.


lexa
Full Member
***
Offline Offline

Activity: 120
Merit: 100


View Profile WWW
July 23, 2014, 03:49:41 PM
Last edit: July 23, 2014, 07:28:53 PM by lexa
 #20

CPU mining:

Code:
minerd-x64-avx --url=stratum+tcp://mrc.eu.blocksolved.com:3310 --user=username.worker --pass=password --algo=scrypt-jane --nfmin=4 --nfmax=18 --starttime=1389028879

Miner download: https://github.com/Thirtybird/cpuminer/releases/

Select right binary for your CPU

Intel Processors
‘avx2’      Intel Core CPU with 64-bit extensions, MOVBE, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AVX2, AES, PCLMUL, FSGSBASE, RDRND, FMA, BMI, BMI2 and F16C instruction set support.
‘avx-i’      Intel Core CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AES, PCLMUL, FSGSBASE, RDRND and F16C instruction set support.
‘avx’      Intel Core i7 CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AES and PCLMUL instruction set support.
‘corei7’   Intel Core i7 CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2 and POPCNT instruction set support.
‘core2’      Intel Core 2 CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3 and SSSE3 instruction set support.
‘nocona’   Improved version of Intel Pentium 4 CPU with 64-bit extensions, MMX, SSE, SSE2 and SSE3 instruction set support.

AMD Processors
‘avx’      Intel Core i7 CPU with 64-bit extensions, MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, SSE4.2, POPCNT, AVX, AES and PCLMUL instruction set support.
‘athlonfx’   Processors based on the AMD K8 core with x86-64 instruction set support, including the AMD Opteron, Athlon 64, and Athlon 64 FX processors. (This supersets MMX, SSE, SSE2, 3DNow!, enhanced 3DNow! and 64-bit instruction set extensions.)
‘nocona’   Improved version of Intel Pentium 4 CPU with 64-bit extensions, MMX, SSE, SSE2 and SSE3 instruction set support.

If you do not know which CPU you have, try each one starting from the top and work your way down.

Pages: [1]
  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!