Topic: Braiins OS & Braiins OS+ custom ASIC firmware: optimize performance & efficiency - page 2. (Read 45390 times)

Hello,
I have Antminer T17 42TH and just install Braiins OS+
( Auto install to NAND from SD card )
But i have an issue with IP address
when i power asic on i can access it for no more then 1 minute
then the ip address is disappear
can not change the ip to static also

You can of course modify the open source bosminer and adapt it to use it the same way they did for cgminer.

Consider the S9 supported there as your base template.

Braiins also develops a closed source version which is used in Braiins OS+, the main difference is the autotuning algorithm and some features such as pre-heating, no hashing when connection is lost, etc.

We don't have any information as to what or when would the open source version get updated. I think if people starts modding it, those changes might actually make it back upstream. Isn't this the spirit of the open source community?

Feel free to add the support of the chips you want, if you distribute it publicly you are required to share back the code changes like other open source projects.

If you are a company or asic manufacturer and have some other business proposals, maybe involving the closed source version, feel free to contact Braiins directly using the form at the bottom of this page:

https://braiins.com/os/plus/partnership

Any plans on supporting things like the GekkoScience Compac F?

Also, is the github project (https://github.com/braiins/braiins/tree/bos-devel/open/bosminer) up to date with the latest development?, seems like it hasn't been updated for 2 years.

So, 27.9 J/TH using factory firmware, vs 22.8 J/TH using Braiins OS+. Is it already reaching the "XP" (next gen) published miner efficiency like it did with the x17?

As for the (estimated) power consumption, like the release notes mentioned, it is still a work in progress. The current version is 2022-03-21 and some improvements have been made, including support for the A series.

any real world data on T17 ? Ideally with upgraded hashboards which should help temps under control/efficiency?

thanks

** Braiins OS+ currently supports: **
Bitmain Antminer S9, S9j, S9i. (No Hydro, K, SE; dual/pro)
Control board C41. (Use slot 6,7,8; no C5, Cyclone or T9+ blue)

Bitmain Antminer S17, S17 Pro, S17+, S17e; T17, T17+, T17e.
Control boards C52, C49. (Use slot 0,1,2; No aftermarket)

** Early access (beta testing): **
Bitmain Antminer S19, S19 Pro, T19, S19j Pro, S19j, S19 A, S19j A, S19j A Pro, S19j A. (No i; +)
Control boards C71, C55, C52, C49. (Use the first 3 slots; No BeagleBone or Amlogic)

** Development pipeline: **
MicroBT Whatsminer M20S, M21S.

** Get notifications from: **
Braiins OS+ web page https://braiins.com/os/plus
Braiins Announcements Telegram channel to: https://t.me/Braiins
Braiins OS+ support Telegram group https://t.me/BraiinsOS

no ... But join my discord if you want test my mod

Does this firmware support S9 Hydro?

User provided results for S19 Pro:



Remember that support for this family of miners is still in "Early access" and work is in progress for further tuning improvements, especially for those not doing immersion.

Which is why you just disable the detection of those fans. The detection works with common fans, but some random brands do that.

Setting a fan to 1% is no reason for the fan to stop talking, but well, at least it didn't go to full RPM like some other random brands do...

This happened to me too.

The problem is that at startup / warmup the software sets the fan speed to a very low value (1% in my case)  and then it complains that the fan is missing when it sees the rpm value drop below a certain (unknown) level. My fan was doing 120 rpm with the 1% command.

I opened a support ticket and was basically told it was all my problem.  That they had seen this with 3rd party fans.

I offered the solution of not checking fan rpm minimum unless the commanded speed level was above, say, 10%. This type of checking is normal for anyone that has done actual real world hardware control.

They said that they may include a "feature" sometime in the future that will allow a user to set some minimum level.

So you just repeat lies like a parrot from "others"...

And yet again, you ignore the facts. Your sole purpose here is to seed FUD is it?
Did i not tell you that there are no added fees at all? YOU seem to have reading difficulties.
So, are you going to ignore the facts? Your lie will stand because you say so? Ask THEM who they deal with and don't involve others in your delusions.

Says the fake account made with the sole purpose of promoting GPL infringement. You must be laughing at Kano as usual...

---

There is Dangerous Temperature, Hot Temperature, and Minimum required fans. Unless you tampered with those or switched temperature control to disabled, this cannot happen. x17 fail on their own when a heatsink loses adhesion due to manufacturing defects (poor quality soldering tin used in the heatsinks), no firmware can save you from that.

As i have said countless times in this thread, Braiins OS can be used straight from the SD card without ever installing it. Read the docs.
It is your sole responsibility to use it, nobody is forcing you to. Installing to NAND is OPTIONAL, again, nobody is forcing you to.

If you installed to NAND, after testing plenty while running from SD, affirming twice your intention but still decide to remove it later, use the bos-toolbox to uninstall. If you messed the NAND by incorrectly trying other sd "recovery" methods, then you have to install Braiins OS from SD again and then use bos-toolbox uninstall.

Yeah it’s dogshit firmware.   Fucking thing couldn’t even figure out it had one fan down and sat there tuning for 3 days and cooked 3 more hash boards.   Fancy interface though if you don’t mind feeding it hash boards!

Did it break your miner too. They should cover our losses they don't care about our hardware just our hashrate.

Anyone know how I can get a S17 back to the factory firmware? 

Slushpool or braiins has a history of stealing from it's customers 

Seems a pretty straight forward question. first line and last line

You having reading difficulty?

Yet you started ranting about it as me spreading lies, to hide from it.

There's no lies there. I repeated things I heard from others about kaboomracks, not from kaboomracks.
So the questions I left quite blatantly open was What say you?

... and yes your site states that it does allow companies to gather fees through using your bos.
But of course I doubt you'd say who does do that, or up until when they did.

Your PM to me saying they don't do it now, doesn't mean they didn't do it in the past either.
I left it up to you to clear it up and you ranted and raved about it.

--

P.S. ranting that you don't like my pool seems a rather amusing thing to do.
As I stated clearly when I changed the N in the PPLNS,
if anyone left and wouldn't stay on the pool, that's their loss for leaving.
But even then it only takes 3 days to get back up to full 'ramp'

Also pretending the other issues I posted never happened is rather amusing

--

Also telling me that I should have done something with firmware ...

Well ... funny you should say that ... you probably should see who it was that made the firmware patches for the S1/S2/S3
... whoops ... that was me

The best known current efficiency records (reported by users) using Braiins OS+ are:

With the S9: 12TH @ 70 J/TH (Bas Laurier)
With the S17+: 53TH @ 29 J/TH (Bas Laurier)
With the S19 Pro: 110TH @ 26.6 J/TH (Jeremy)*
All cases measured at the wall.

Efficiency, is what produces more long term profits, not brute speed. Its not just electricity price, it is the environmental managing (noise/heat) and the increase of downtime from stressing more the components.

The reason why an S19 is better than an S17, or an M30S better than a M20S, is that, efficiency. The amount of watts each TH/s needs. And the more machines you have, the more it weights.

Those who go crazy with speed, end losing. If they overclock their S19, making them perform worse than the S17, then anyone could by buying two S17 and making them efficient, produce more TH/s, use less electricity and pay less in machine costs (because 2 S17 cost less than 1 S19).

The excuse that "my electricity is free" is invalid, because even if they may not pay (which by the way would be very dubious, someone pays, always) this isn't going to save them from not being able to make use of the full capacity of the load to the max, or from frequent downtime due to maintenance and repairs.

"Less (power) is more (profit)".

*S19 family support is in early access and is a work in progress, expect further improvements in later versions.

Let me remind you that this is the Braiins OS thread, not the Slush Pool thread. I'm not a mod, and unlike your thread, this is not self moderated.

So, repeating the same old story of 2015 that nobody cares about, that was fixed back then. Fact is the pool works and unlike yours, its still commercially viable to use it. Nobody is going to wait a year to get paid, you are solely responsible for this, by failing to act when you had the chance and the pool was still viable. The irony is you let in the factory infringing firmware but blocked legitimate software, well its not like you needed the hashrate anyway...

You and your perfect pool that nobody wants to use. By mismanaging it, you owe the community even more, all the people that simply had to leave after waiting long enough. The ruin of kanopool simply left the community without a viable pplns pool that is not invite only like those whale American pools. Everybody else is PPS or worse (multi coin), only Slush Pool stands alone to accommodate both small and large miners and NOT do PPS.

What has that old story about Slush Pool to do with Braiins OS, the mining firmware that can be used in ANY pool? Nothing.

Whats the point of reading bosminer code? Well, you could learn about the proposal you keep bashing, and what makes it different to other miner software. Then you would learn how the communication works, perhaps even implement it (better?). The changes in the newer versions you probably don't need, since you already know how to talk with those chips. You might as well add the x17 support yourself no? Of course not, you have cgminer...

So what if its also closed source? Is dual licensing forbidden? No, would you rather support infringing cgminer? Seems you do, because Braiins, unlike your former Chinese employer, has not ever infringed the cgminer's license, even back when it was used for a couple of years.

By 2020 Braiins pushed ahead and did the right thing: Replace it. Its Braiins property now, and Braiins can decide the terms of use however Braiins sees fit. Leaving this component aside, the rest of the OS is OpenWRT, not that Xillinx half made nonsense Bitmain (and modders) use. Even MicroBT uses it, and a gazillion embedded devices. It is a more proper distro, even you know that. Braiins forked it and called it Braiins OS adding some minor cosmetic changes and its own miner software.

Kano could have done it, but Kano didn't. All Kano does is bash others for no reason, including the creator of cgminer, as much as you say you coded more, its still the project from someone else. You will never escape this fact, since you didn't want to rename your fork and let cgminer rest in peace with ckpool (software), etc. To add confusion, now people say "Kano's cgminer", well whatever, i told you kanominer made more sense...

Kano says V2 is bloat, but Kano has not proposed a solution to end the clear text situation in V1. Even if you ignore the other features (they are optional, after all) this is already a reality in the current V2 as implemented by bosminer and Slush Pool. The Chinese did something as well but did not spread. Stratum with its clear text json messages keep revealing people's private information to others 24/7. But you don't seem to care since you live in a nice country.

The lie is that nonsense you also repeated elsewhere about a "9% hidden fee" involving an ASIC reseller, why don't you read your own messages? Its right here in this thread and in your pm. Why don't YOU ask THEM instead what is what they use?

Kano read your messages.

What point is there downloading the miner that is not what you are promoting and wanting everyone to run?
That's the old open source version.
My factual comments are about your current closed source miner that you bandy to S17 and above.


Why wouldn't anyone ignore it? The majority of it is basically Luke's GBT that died and no one implemented a transaction selection for.
WAY too much data.
Needs a WAY too fast computer to run each miner to implement transaction selection.
But it wasn't implemented for the blatantly simple reason that it creates bias in Bitcoin.
Transaction selection should be based on available transactions, not some biased view about which transactions should and shouldn't be allowed.
You clearly misunderstand the point of Bitcoin.


Um - what's all this crap you stated here?
To avoid commenting on or replying to what I actually asked? ...

It's what did happen on multiple occasions - here's 3:

Firstly quoted from your web site:
"Braiins has been the sole operator of the pool since taking a majority stake in 2013."

So all the below are you.

1) Dec 2015 - during that month, miners on your pool posted issues with the luck during the month.
Your support said there were no problems.
After that month I did a valid and correct statistical analysis of the slush blocks for the month:
https://bitcointalksearch.org/topic/m.13482822
This CLEARLY shows there were problems that slush allowed, in it's negligence, to continue well past the point of obvious detection.
This also CLEARLY shows that either 1) for the first 5 years you and slush ran the pool, you had no useful statistical analysis of the pool to handle accidental (or purposeful) block withholding OR 2) you didn't bother to take note of any of that analysis that would have identified the problem around half way or less through the month.
Your pool result: all miners lost well over 10% of their earnings for that whole month - more than 350 BTC
Your comment: 'problem is fixed' - implying quite clearly: bad luck, you lose, due to slush negligence.
Followup, it was almost certainly organofcorti who told you how to do the so called 'proof' you now supply about your pool hash rate.
Alas, you ignore the fact that it is not a proof of hash rate, it is a proof of a lower limit on the hash rate,
which for a PPLNS pool is quite a problematic claim.
Of interest to also note that organofcorti pointed out a method that would have identified the above problem on the first day of the month.
https://bitcointalksearch.org/topic/m.13892442
I wonder if you do this every single day to mitigate such losses for your miners

2) Jul 2018 - two related events:
Slush: 5 blocks diff%  1117%, 127%, 237%, 119%, 509% - a one in 717839 blocks event.
https://bitcointalksearch.org/topic/m.41534985
The year before, my pool had it's worst ever bad diff run, that no matter how I organised the 5 blocks, it wasn't as bad as the slush luck.
My response was to code a 2nd original share and block test, and reprocess and re-verify the 3 months worth of shares around the event.
Reported all this during and after the event and the results, suggesting it was either very bad luck or FuckHash mining data that was the largest hash rate on the pool at the time. PPLNS hasn't allowed rentals ever since due to the risk issues with rentals.
Meanwhile Slush Official's response to their bad luck above:
https://bitcointalksearch.org/topic/m.41904145
"Bad luck happens"

3) However, during that 5 block ridiculous bad luck run, there was a new miner on slush pool with 400PH which equated to about 2 blocks a day at the time.
They had no blocks in the "Hall of Fame" for that period nor the week after it.
Slush Official's response to the above:
https://bitcointalksearch.org/topic/m.43143004
"Hall of Fame is rather "for fun" statistics"
This seems a rather major issue to claim this on a PPLNS pool where reward is determined by blocks found.
This relates directly to the fact that slush's "proof of hashrate" is not exactly that, it is a proof of a lower limit.
A higher hash rate can easily be hidden by using a different coinbase and thus it not showing up in the proof.
Worse, when a miner that is indeed failing to find blocks, it will mean that the pool 'proof' will indeed show this lower limit.

4 Then of course there's the blatant lie you continue to advertise on your home web page claiming you were the first mining pool, even though you clearly know it is false, and even slush himself knew it since he was involved in the thread where he pointed out when he started slush pool that there was already another (in the same thread).

Where exactly did I say this?
All you've done is run your mouth to avoid some of the questions I asked.

Where is the lie?
Seriously - point it out.

You still owe over 350BTC to the slush mining community Dec 2015 for your negligence in one incident above

---

Pretending ignorance doesn't allow to you ignore the license.

You distribute a full firmware.

Thus you distribute cgminer in your firmware.

This requires, according to the GPLv3 License, for you to provide the source upon request.

[moderator's note: consecutive posts merged]