Topic: [Mining OS] mmpOS - easy to use linux/windows platform for mining. - page 3. (Read 5132 times)

We have added Folding @ Home as "miner", it is intended for fighting with COVID-19 decease by providing compute workloads for solving potential vaccine https://foldingathome.org/2020/02/27/foldinghome-takes-up-the-fight-against-covid-19-2019-ncov/. When selected as coin[Folding at Home], set wallet name to none or something made up, same goes for pool, set hostname to non existing/non valid, it is not taken. If you have FaH passkey you can add it in advanced miner arguments[Miner profile] --passkey=yourpasskey .
!!! Warning !!!
All samples are downloaded in tmpfs so on reboots they do not persist, if you want to make it permanent - do not use usb drives and edit /opt/mmp/miners/foldingathome/mmp-stats.sh replacing storage="/run" with storage="/home" or some directory in non tmpfs.

This is not added for profit, so questions like "How do I get paid?" have simple answer - "You don't". You donate your compute resources for fighting not only with COVID-19 but other deadly deceases.
We will do everything we can to assist everyone who is willing to donate compute resources.
Stay safe, stay healthy, stay home and be responsible!

Feel free to join us on telegram or discord for more information.
https://discord.gg/P5J6Z5h and https://t.me/mmpos_intl

# 29. February 2020 - System driver upgrades

* Updated linux kernel to 5.4.22 in unstable.
* Updated AMD driver to 5.4.4 in unstable variant.
* Updated NVIDIA driver to 440.64 in unstable variant.
* Updated linux kernel to 4.19.101 in stable variant.
* Updated AMD driver to 5.2.7 in stable variant.
* Updated NVIDIA driver to 440.59 in stable variant.
* Added timeout during rig registered check for faster startup.
* Introduced per GEO Location mirrors for updating drivers.
* Introduced per GEO Location apt repository mirrors.

# 12. February 2020 Update agent

This release improves Overclocking and Undervolting of AMD Navi cards and also added fan control for all aftermarket AMD Navi cards.

* Updated agent to 2.8.7
* Added AMD Navi10/12/14 memory voltage control
* Updated AMD Navi10/12/14 fan control, should now work on all cards
* Improved memory detection for all AMD cards from Vega and above
* Improved smart tune of supported NVIDIA cards and AMD Vega56 with Hynix memory
* Added detection of AMD RX 5600XT and RX 5500XT cards

12. February 2020
* * * Image and System Upgrades * * *
* Updated linux kernel to 4.19.98 for stable variant.
* Updated linux kernel to 4.19.101 for unstable variant.
* Updated AMD driver to 5.2.6 for stable variant.
* Updated NVIDIA driver to 440.44 for stable variant.
* Updated NVIDIA driver to 440.59 for unstable variant.
* Updated base to use lsb-release for future mmpOS 3.0 integration.

Stay tuned for today's agent update

# 31. January 2020 - Dashboard update

* Added long awaited referral system!
* Refer your friends and you both get an additional 10% bonus credits on every credit purchase your referee makes.
* Referral system applies to already registered members and is limited to only inability of changing your already entered referral code and you can't refer yourself

# 07. January 2020 - AMD Driver update

* Updated AMD driver to 5.2.5 in stable variant.
* Updated AMD driver to 5.2.6 in unstable variant.
* Increased stability on voltage starvation for Vega10, Vega20 and Navi ASICs.
* Fixed Hawaii powerplay warning in some cases.
* Fixed Polaris memory allocation in unstable variant ( May cause Polaris GPU to get stuck in weird cases ).

I can assure you it makes no difference in linux.

Thank you. I managed to expand. I also tried "swap --add" but it seems it only creates swap file of 524.284 MB, is there a way to increase it? I would like to test if it makes any difference.

Thanks.

Hello,
Linux does not need virtual swap space like windows. Regarding expanding its in ‘mmp help’ list - mmp expand via ssh.

Hi,

I flashed this into a 120GB SSD using Rufus and ended up with only about 10GB used and the other 110GB being un-allocated. Is there a way expand the OS into the remaining space?
I am looking to mine ETH on my Vegas and my understanding is that it benefits immensely from having large pages, hence why using the extra space would be nice.

Thank you.

# 23. December 2019 - MMP Stable release

* MMP 1 year anniversary update.
* We are releasing our first stable release, no longer public beta!
* Ubuntu 18.04-LTS based image will still be tagged as beta and is intended to replace our 16.04-LTS stable one.
* Added support for RX5700/5500(Navi ASICs) in stable variant.
* Updated kernel to 4.19.91 LTS.
* Added driver for BCU1525 in mmp-addin-drivers for enthusiasts.
* Updated Legacy NVIDIA driver to 390.132
* Added more checks for dpkg recovery.

# 18. December 2019 Update agent and dashboard

With this release we are rolling out our first iteration of redesigning the dashboard.

Our next plans are combining the rig info dialog and configuration and also making the Overclocking presets more accessable.

* Updated agent to 2.8.6
* Added smart tune for compute heavy algos on recent AMD CPUs
* Collecting and showing core + memory usage of GPUs and CPUs in dashboard
* WIN: Showing ASIC+Memory Temperature of AMD Vega+ cards on windows

# 11. December 2019 Update agent and dashboard

* Updated agent to 2.8.5
* Added smart tune for compute heavy algos on recent Intel CPUs (up to 30% increased performance on med/high)
* MMPOS: fixed powerlimit settings for AMD Vega56/64/FE
* MMPOS: improved watchdog to detect more AMD related issues
* MMPOS: added option to control HugePages settings via dashboard

AMD Vega56/64 Powersave recommendations

GPU | COIN | MEMORY VENDOR | CORE CLK | MEMORY CLK | POWERLIMIT | VOLTAGE | TUNE
RX Vega 56 | ZCOIN | SAMSUNG | 1220 | - | - | 0.818 | Default/Compute Heavy
RX Vega 56 | ZCOIN | HYNIX | 1220 | - | - | 0.818 | Default/Compute Heavy
RX Vega 64 | ZCOIN | SAMSUNG | 1220 | - | - | 0.818 | Default/Compute Heavy

RX Vega 56 | GRIN31| SAMSUNG | 1220 | 900 | - | 0.825 | Default/Cuckoo*
RX Vega 56 | GRIN31 | HYNIX | 1220 | 900 | - | 0.825 | Default/Cuckoo*
RX Vega 64 | GRIN31 | SAMSUNG | 1220 | 1025 | - | 0.832 | Default/Cuckoo*

RX Vega 56 | ETHASH | SAMSUNG | 965 | 960 | - | 0.818 | Default/Ethash*
RX Vega 56 | ETHASH | HYNIX | 950 | 940 | - | 0.818 | Default/Ethash*
RX Vega 64 | ETHASH | SAMSUNG | 1023 | 1075 | - | 0.850 | Default/Ethash*

RVN and RYO Low power.
RX Vega 56 | RVN/RYO | SAMSUNG | 1220 | - | - | 0.818 | High/Compute Heavy
RX Vega 56 | RVN/RYO | HYNIX | 1220 | - | - | 0.818 | High/Compute Heavy
RX Vega 64 | RVN/RYO | SAMSUNG | 1220 | - | - | 0.818 | High/Compute Heavy

RVN and RYO high performance.
RX Vega 56 | RVN/RYO | SAMSUNG | 1450 | - | - | 0.868 | High/Compute Heavy
RX Vega 56 | RVN/RYO | HYNIX | 1450 | - | - | 0.868 | High/Compute Heavy
RX Vega 64 | RVN/RYO | SAMSUNG | 1490 | - | - | 0.875 | High/Compute Heavy

*Note that you can always test medium and high tune factors on ethash and cuckoo agorithms, but for MTP we do not recommend using medium/high preset on Compute Heavy algorithm


Recommended miners for popular coins per platform:

NVIDIA:
Ethereum(ETH+ETC)  and forks -> ethminer by ethereum mining community(recommended by us), TT-Miner and gminer
GRIN29 -> gminer, nbminer and bminer
GRIN31 -> gminer, nbminer and bminer
BEAM -> gminer and miniZ
RVN -> t-rex and zenemy
ZCOIN -> TT-Miner(recommended by us), ccminer-mtp and t-rex


AMD:
Ethereum(ETH+ETC) and forks -> teamredminer(recommended by us) and ethminer
Cryptonote coins -> teamredminer(recommended by us), xmrig and xmr-stak.
GRIN29 -> teamredminer and lolMiner(both recommended by us)
GRIN31 -> teamredminer and lolMiner(both recommended by us)
BEAM -> lolMiner(recommended by us)
RVN -> teamredminer(recommended by us) and wildrig-multi
ZCOIN -> teamredminer(recommended by us), sgminer-fancyX and wildrig-multi

CPU:
Only suitable algorithm for now is RandomX which is currently supported on our platform by xmrig and xmr-stak-rx.
We support cpuminer-opt which has too many coins support but none of them is profitable by only CPU mining.

Disclamer: Some of recommended miners by us may not be the fastest but the developers behind them had proven trustworthy and we support them and their work.
Our personal recommendation should not be considered mandatory! You are free to use whatever miner you find best for your needs!


If you have found something better or more efficient, or performance wise, please do share your settings with us, we will really appreciate it!
Thank you!

# 30. November 2019 Update agent

With this release we introduce a completely new mining experience for your AMD RX Vega56/64 GPUs.

We created a special smart tune algorithm option called *compute heavy*, which reduces power usage without hashrate penalty for compute heavy algos like MTP, CNGPU, x16r/t/v2, x21, x22i, x25x and Lyra2RE.

Everything we do is driven towards success of our not so big community. We have put a lot of effort creating these features, so users save power and enjoy profitable mining again.
So, MMP OS community, **Thank you for staying with us and believing in us!** We will do our best to keep the pace in MMP OS development process.

We would like to say **Thank you** to teamredminer (https://github.com/todxx/teamredminer) developers(todxx and Kerney666) and their awesome community for creating their amazing miner and helping us become better!

We would also like to say **Thank you** to lolMiner (https://github.com/Lolliedieb/lolMiner-releases) developer Lolliedieb for his outstanding work and contribution to the mining community.

* Updated agent to 2.8.4
* Improved memory based smart tune for AMD and Nvidia
* MMPOS: added compute heavy smart tune for AMD Vega56/64
* MMPOS: optimized power usage for AMD Vega56/64

We already do. A few hours before the fork we will switch arguments line in advanced miner settings for xmrig from cn/r to rx/0, same goes for xmr-stak-rx.
RandomX was introduced with our CPU mining support a few weeks ago. You are welcome in our discord/telegram channels if you have any problems setting it up.

are you guys gonna support randomx?

# 08. November 2019 Update agent and experimental variant

This update makes AMD Navi cards fully supported if you use the experimental variant of mmpOS.
Navi is still not supported by all miners but you are now able to mine and overclock if you have compatible miner and switch to the experimental variant.

* Update agent to 2.8.3
* Support for Overclocking and voltage control of AMD Navi(RX 5500/5700) cards.

# 07. November 2019 Update agent

* Update agent to 2.8.2
* Optimized miner download verification (ensure hashfile is valid)
* Improved startup with unsupported GPUs, we now allow startup after a 60second retry phase even if detected pci devices do not match driver reported ones.

# 02. November 2019 Update agent + dashboard

This update introduces CPU mining. Similiar to multi-miner support, just add a cpu profile and start mining.
As with GPUs we provide metrics like Temperature and Voltage for recent CPUs as well.

* Updated agent to 2.8.1
* Added CPU worker support to agent and dashboard
* Added native xmrig protocol support
* Optimized AMD memory detection. (Navi and others are supported too)
* Fixed Hawaii and Fiji memory overclocking bug

# 25. September 2019 Update agent + dashboard

This update is a big one, we now support running multiple miner instances per platform (Nvidia, AMD and SQRL Acorn's).
So currently you can run up to 3 miners in parallel.

If you mine with multiple miner instances to the same pool, we recommend to add %miner_id% to your pool username.
It will suffix the username if miner instance is > 0 (so first miner will be rigname, second one rigname1)

Next step will be supporting CPU mining.

* Updated agent to 2.8.0
* Added multi-miner support to dashboard and agent
* Agent no longer queries fpgas for info. Every stat will now be gathered via miner.
* Added native ninjarig protocol support
* Added multiple new amd gpu and memory types to detection
* WIN: Now also supporting the "automatic mode" to add a new rig
* WIN: Added voltage display support for recent nvidia cards