Braiins vs Luxor - Error Rates and Rewards

I opening this thread as I’ve started a comparison between these two pools.


Subbed! Looking forward to your first data dump on this

We had a planned outage for some grid maintenance about a day ago. The miners, obviously, rebooted. Luxor is killing it in regards to HW errors. Braiins, quite a bit different.

The middle two miners are on braiins. The outer two are on luxor. Again, these are identical miners and historically they all perform about the same and have about the same HW error rate when on the same pool.

I’ll post up again near the end of the month unless there is some other reason to do so before then.

My main focus here is if Braiins can perform similarly to luxor then it would be the obvious choice for Braiins OS (free basically). I would be highly interested in error rates on a “tuned” miner. If braiins does not perform to my liking then perhaps the ePIC UMC board and firmware will be the way to go ($150-160 a miner). I would also be more apt to go with the ePIC UMC board if it is compatible with the new S21 series antminers. Another aspect that I didn’t really think of is power consumption between pools. I hope to get into some circuit monitoring at some point this year.

The deciding factor for me to not continue mining BTC with Luxor Pool is that they charge transaction fees on their payouts. At one point when transaction fees were high last month, I had to pay 4%-5% of my payout for the transaction fees.

Where are you viewing this information? @faheshang

This was stated in their docs page last month:

Many miners were complaining on their discord channel about Luxor not
paying transaction fees on payouts.

After 4 emails to them asking about their withdrawal fees policy they sent me this.

That’s definitely something to consider. Thank you very much for providing this information. I’m not providing any data yet, but preliminary info is showing braiins and luxor as pretty close on payouts. Errors rates have mostly leveled out as well between both.

I do have a current monitoring device that will be here this week as well. I will take some readings on both pools as well to see if there is any noticeable difference. I doubt it, but something to look at.

I’m not sure which way I’ll go from here since it is so close.


date reward_total
2024-01-14 0.00057016
2024-01-15 0.00051953
2024-01-16 0.00051186
2024-01-17 0.000515
2024-01-18 0.00051342
2024-01-19 0.00050049
2024-01-20 0.00049717
2024-01-21 0.00050668
2024-01-22 0.0005064
2024-01-23 0.00051333
2024-01-24 0.00048827
2024-01-25 0.00049645
2024-01-26 0.00050022
2024-01-27 0.00049403
2024-01-28 0.00049479
2024-01-29 0.00049583
Total 0.00812363


date revenue (BTC)
01/29/2024 0.0005003
01/28/2024 0.00048846
01/27/2024 0.00049377
01/26/2024 0.00048579
01/25/2024 0.00049241
01/24/2024 0.00049827
01/23/2024 0.00050269
01/22/2024 0.00050242
01/21/2024 0.00050763
01/20/2024 0.00049727
01/19/2024 0.00049498
01/18/2024 0.00050461
01/17/2024 0.00052309
01/16/2024 0.00051888
01/15/2024 0.00055405
01/14/2024 0.00053587
Total 0.00810049

0.2% difference between the two pools. I mean you can’t really get any closer than that.

The one thing that keeps nagging me is the error rate.

The two outer miners (1 and 4) are on luxor. The two inner miners are on braiins (2 and 3). I notice that with braiins I seem to have higher errors numbers initially and then over time they slowly drop, whereas luxor seems to start off low and mostly stays lows.

I may move them all back to luxor and see if the power consumption changes on the overall setup at all and to also check the error rates on the middle two. I’ll post back any findings.

edit: just moved all back to luxor to initiate my test. I’ll probably let those run for a week to let them settle out in regards to errors.


After a couple days or so this is where we are at. The outer ones have been up for 1.2 weeks and the two inner miners have been up 2.43 days. #2 did start off with a couple errors that made it work down to where it is at now, but I think it shows it is fairly obvious there is a difference in rates.

With that said, I have braiins’ toolbox setup and may get some custom FW installed soon.

@VoskCoin How are the S19XP units you have doing immersed and with braiins? Do you find they ever restart/reboot? I’m mostly after stability. I’m not sure I will be able to collect data right off the bat without having to modify the script that scrapes all this data off of the machines.

1 Like

Installed brains. Massive issues. I reverted back to stock. Hopefully nothing damaged. This was my concern.

Edit: I will say the firmware did install and revert very easily. I’ve been up again for 40 minutes with seemingly normal operation on stock firmware. Fingers crossed. I was able to get the logs quickly out of the BOSminer before reverting. I have forwarded them to Braiins to see what they say. The following is happening on every hashboard constantly.

WARN bosminer_backend::metrics: Event “crc_uart_rx” on hashboard 2 happend 50.8 times a second (limit is 4.0)

In the stock firmware logs you will see a reg_crc_error every once in a while, so I’m not sure what is causing this. Hopefully braiins has some insight.

1 Like

The only thing I don’t like about braiins,, Vnish is because if you look at the power consumption (on the wall not dashboard) it’s ridiculously high during the auto tuning process. This might not be a problem trying to underclock your miners, but when I tried to overclock my J pro to 120T it started the auto tuning process at 4800 watts and held at that for over an hour like that, and it did it for all 3 of those firmwares. That’s the only reason why I like Luxos is because it starts slow, ramps up and warms up instead of been thrown into the deep end from the get go

@ShredZ Does LuxOS have a way to install on AMLogic boards?

Yes they now have a thing called commander so you can do mass installations on your network, however when I asked for a license key I keep getting an error for invalid license key even though they gave me 3 keys, but everyone it works for says it’s easy. However if you wanted to remove the firmware from amlogic im not sure how that works, perhaps email their team first so if you decide you don’t like it you know how to get it off (and ask for a license key when you email them)

@ShredZ Do do they remove fees if mining on their pool? I suppose I could look it up too. ha

Yeah I think either the dev fee or pool fee is scrapped

I have been noticing higher rejection rates at Luxor lately for some reason. I’m not sure if I’m collecting that data, but let me take a look.

Nope, haven’t been collecting that information. Not sure if it is available through CG.

I did look it up and I have it setup to be able to collect that info. I will just need to spend an hour or so to finish setting it up to actually collect it. I’m not sure when I’ll have that time, but I will probably add it in soon.

edit: oddly luxor is late on rewards for the day. I’m wondering if the “rejection” rates I’m seeing is not artificial and they’re having a hard time paying out since transaction fees have been so low. It just seems odd they’re floating around 1% or more all of the sudden.

I heard back from Braiins support and they said there was nothing wrong and I needed to lower the power limit 100W. That doesn’t make much sense since it was set at the stock power output of 2881W. I’m a bit hesitant to load the firmware again with that experience.

Edit: I did just reach out to see if I can disable the autotuning and if I do if it will run identically as the stock firmware in the sense of how it would “tune”.

I just attempted the firmware again with lower power target settings and the same issue existed. Thousands of these errors:

WARN bosminer_backend::metrics: Event “crc_uart_rx” on hashboard 2 happend 50.8 times a second (limit is 4.0)

It does it across each hashboard, so it’s more or less either something in the software or an issue with the miner(s) that only happens with the way their software works.