The bitaxeSupra: Open source Bitcoin miner based on the BM1368 ASIC

[cygan]

a new baby is born!
Skot presented the new bitaxeSupra today with the following tweet. it uses the bm1368 asic chip, which can also be found in the Antminer s21
a hash rate of about 620 gh/s should be achievable, the picture even shows a hash rate of over 700 gh/s...

my question is, since the bitaxeHex is not yet finished, which chips will be used there, the 1366 or the 1368?


https://twitter.com/skot9000/status/1755262407197303119

[1714283124]

1714283124

[1714283124]

1714283124

[1714283124]

1714283124

[kekeck33]

Can you buy BM 1368 chip?

[cygan]

Can you buy BM 1368 chip?

according to Skot, since the chips are relatively new, they are only available in sets of 324...

[BitcoinSoloMiner]

I want a hex with these. 19W/Th... Nice!

[kekeck33]

I want a hex with these. 19W/Th... Nice!

Me too

[Gholly]

Did they just use the ultra hardware and substitute the BM1368 chip?
Do they have the same pin-out?
I don't see anything on the github or bitaxe.org about this model..

[stompix]

according to Skot, since the chips are relatively new, they are only available in sets of 324...

 
and not only that:

You've got to protect those beauties in shipping by soldering them to a PCB and encasing them in heatsinks

Probably this is where he got it:
https://twitter.com/skot9000/status/1743057204926882303

[mikeywith]

Can you buy BM 1368 chip?

according to Skot, since the chips are relatively new, they are only available in sets of 324...

Zeusbtc has them with 10 MOQ, there is no price tag but I think it's in the 40$ range, in a few days the Chinese New Year is finished, if anyone is interested I can confirm the price.

and not only that:

You've got to protect those beauties in shipping by soldering them to a PCB and encasing them in heatsinks

Who made the joke? 

[stompix]

Who made the joke? 

It was skot9000 himself:
https://twitter.com/skot9000/status/1755326025943163030
and he seems like he didn't get it from Zeus either but directly from Bitmain: (from his tweets)

nobody is selling individual chips yet. you have to buy a "pack" of 324 chips attached to this pesky shipping container called a S21

But at 40$ it would be a bit expensive but not that prohibitive for hobbies and stuff.
If you buy the whole miner for 5k you get the chip at 15$, I don't know how much the other stuff is but you could probably glue it together for another 30, rounding it to $70, so of course it won't compete in ROI with the S21 but a million of those could compete in profitability, no hobbyist is going to care about a 10w chip running on the desk.
Either way this would be the only way to true decentralization.

[philipma1957]

Who made the joke? 

It was skot9000 himself:
https://twitter.com/skot9000/status/1755326025943163030
and he seems like he didn't get it from Zeus either but directly from Bitmain: (from his tweets)

nobody is selling individual chips yet. you have to buy a "pack" of 324 chips attached to this pesky shipping container called a S21

But at 40$ it would be a bit expensive but not that prohibitive for hobbies and stuff.
If you buy the whole miner for 5k you get the chip at 15$, I don't know how much the other stuff is but you could probably glue it together for another 30, rounding it to $70, so of course it won't compete in ROI with the S21 but a million of those could compete in profitability, no hobbyist is going to care about a 10w chip running on the desk.
Either way this would be the only way to true decentralization.

but you would need millions of 10 watt devices. as you would want the product to have 1%-2% of the hash rate.

also would need many nodes.

[stompix]

but you would need millions of 10 watt devices. as you would want the product to have 1%-2% of the hash rate.
also would need many nodes.

Yup, millions and even hundred of millions, as to come closer to the hashrate right now you would need over 800 millions chips ...
But if we manage to get more of those and not just at 10W, maybe even higher with 50W, and as their hashrate will start making larger farms income go down you could probably achieve some level of true decentralization with fewer devices.
Imagine how big farms would feel with the halving coming and knowing half of the hashrate is run as fancy wallclocks and their owners don't give a damn about the price per kwh?   

[philipma1957]

but you would need millions of 10 watt devices. as you would want the product to have 1%-2% of the hash rate.
also would need many nodes.

Yup, millions and even hundred of millions, as to come closer to the hashrate right now you would need over 800 millions chips ...
But if we manage to get more of those and not just at 10W, maybe even higher with 50W, and as their hashrate will start making larger farms income go down you could probably achieve some level of true decentralization with fewer devices.
Imagine how big farms would feel with the halving coming and knowing half of the hashrate is run as fancy wallclocks and their owners don't give a damn about the price per kwh?   

Exactly a 25 watt node with a chip or 2 would be best. But price needs to be low.  there are a few units that are close to what would be best.

This unit is good very efficient, it could use 2 chips vs 1 and a node option.  At a good price.

The new Apollo is 4th with a node but is not efficient as this and is far too costly.

Canaan has the nano it is not as efficient and it does not have a node option but it is cheap.

If you could build a 2 chip 19watt 1th miner with a node and sell it for about 400 with the node and 150 without the node it would work.

 But I do not see this coming any time soon.

It would make it decentralized for sure.

[mikeywith]

It was skot9000 himself:

It was an obvious joke, he meant you need to buy the whole miner from Bitmain which comes with 324 chips, chips don't need that much of protection.

If you buy the whole miner for 5k you get the chip at 15$, I don't know how much the other stuff is but you could probably glue it together for another 30, rounding it to $70, so of course it won't compete in ROI with the S21 but a million of those could compete in profitability, no hobbyist is going to care about a 10w chip running on the desk.
Either way this would be the only way to true decentralization.

The cost per hash would be a lot higher anyway, you can only put so much in a single non-Bitmain miner, but would require the same controller and logic, I honestly see no point in these small miners, all of them will be used for lottery mining, and there is no such thing as "improve decentralization" with these expensive small miners, if you are not finding blocks -- you technically don't exist, hundreds of thousands of people need to operate these tiny miners to find a few blocks that don't even count.

The best thing we can hope for in terms of decentralization is that another country besides the U.S starts mining BTC heavily and takes 20-30% of the hashrate, until that happens, it's all fudge.

[stompix]

The cost per hash would be a lot higher anyway, you can only put so much in a single non-Bitmain miner, but would require the same controller and logic, I honestly see no point in these small miners, all of them will be used for lottery mining, and there is no such thing as "improve decentralization" with these expensive small miners, if you are not finding blocks -- you technically don't exist, hundreds of thousands of people need to operate these tiny miners to find a few blocks that don't even count.

I did said that's the only way to true decentralization, not that it could be done....
Theoretically having millions of those devices and 10-20 pools with the devices having an algorithm that randomly changes the pool every every now and then and with people not caring about how much that things consume and how much it earns you could achieve something decentralized like a like a torrent distribution network where everyone even with 10gb and 10mbit is contributing to the network.

This system would still have a cost advantage, after the initial purchase users won't care that much about the cents/kwh, there is no maintenance guys, there is no rent no cooling, no wages to employees, but again that's theory on how could ....bla bla bla...

The best thing we can hope for in terms of decentralization is that another country besides the U.S starts mining BTC heavily and takes 20-30% of the hashrate, until that happens, it's all fudge.

And as the hashrate grows and so does the need of electricity, how many countries out there could host it?
If we say 1 million s21 that's at least 3500MW, assuming the country would need for that just 10% extra of its total generation and you're down to 16 countries.
There are only 3 realistic possibilities, Canada (or the US but with polar bears and maple syrup), Russia and I really don't think that having that country host  an important share of the hashrate would be that much of a great idea, and Brazil, probably the only realistic choice. But...nobody is moving there!

[mikeywith]

Theoretically having millions of those devices and 10-20 pools with the devices having an algorithm that randomly changes the pool every every now and then and with people not caring about how much that things consume and how much it earns.

Ya but that is a lot of "ifs" which we all know isn't going to happen, there are not enough people (millions) who can afford and care about "mining decentralization" without caring about the cost/profit, also given that we talking $100-200 per th, it just not going to happen, even if you had millions, speaking of the device in the this very topic, it does 0.7th per device? so you are going to need 1.4M of them to obtain 1EH which is not even 0.2% of the total hashrate, so what are the numbers we talking about? we want these small miners to gain 25% of the network hashrate? great then we need 210,000,000 of them.

I know what you are saying is "in theory", I am just saying it's not even possible in theory.

you could achieve something decentralized like a like a torrent distribution network where everyone even with 10gb and 10mbit is contributing to the network.

Torrent is built on a different structure, anything above 0 data and bandwidth will benefit the network, even if it's a 1kb txtfile and an old dial up connection you would actually contribute to the overall streght and availability of the P2P network, in BTC, it's different, if you are not finding blocks -- you are just burning electcity for no good reason.

And as the hashrate grows and so does the need of electricity, how many countries out there could host it?

I believe the opposite will be true, the more times goes buy, the less eleciticty BTC would use, of course, we are not there yet, but I think of it as climbing a mountain, we would get to the top of it and then start going down, the only reason why the market still allowes electcity growth is because we have not hit the effiecny wall and the rewards are still huge, eventually, when there is a little room left for hardware improvement, and the block rewards are next to nothing, there bitcoin ecosystem won't have a room large for everyone.

Obivouslly, given that we don't know which gears are running at the moment, we could have already been in that stage, talk for an instance the S9 era where hashare was in the 60EH, it has grown by a factor of 10 today, obviously, the power consumption did not, and if we were to assume that it was only S9s then (the most effiecnt at the time) and assume that it's all S21 or M60s today, then effiency has improved by 80%.

60EH back then needed 6,000MW (if all are S9s)
600eH today needs 12,000MW (if all are latest gen miners)

So while we increased the hashrate by 900%, we increased the consumption by only 100%.

Obviously, not all gears today are 20w/th, also not all gears back then were 100w/th, these numbers are somewhat random to explain my point of how in the future we would likely be able to achieve power demand decrease.

Russia and I really don't think that having that country host  an important share of the hashrate would be that much of a great idea

Is that a politically biased opinion or there exists a valid technical reason behind it?

But regardless of what you think about said country, I doubt you think they are worse than China, do you? the Chinese had control over bitcoin mining, and nothing happened, personally, I don't believe "evil" would exist in the mining business regardless of the government views of BTC, I highly believe that rich people are stronger than all governments, if there exists a country where a few rich people own 20-30% of the total hashrate, it means they are strong enough for any government to screw them over, I know what some legacy media might tell you about countries that are ruled by a single man -- they say the same thing about my country, but I know such things don't exists, power resides where money resides, the government (or otherwise a single handed man) is only strong enough to control the average citizens, the rich -- will always operate as they please and will always do what's best for their investment.

So ya, we could say x country is bad and y country would be a better option for BTC mining, to me, I don't care what the second country is, even if it was in a different planet, it's certainly better than all of it being in one place, but again, as we know, people from different countries/planets can always come together, sit behind closed doors and do whatever the fudge they want to do.

[philipma1957]

Theoretically having millions of those devices and 10-20 pools with the devices having an algorithm that randomly changes the pool every every now and then and with people not caring about how much that things consume and how much it earns.

Ya but that is a lot of "ifs" which we all know isn't going to happen, there are not enough people (millions) who can afford and care about "mining decentralization" without caring about the cost/profit, also given that we talking $100-200 per th, it just not going to happen, even if you had millions, speaking of the device in the this very topic, it does 0.7th per device? so you are going to need 1.4M of them to obtain 1EH which is not even 0.2% of the total hashrate, so what are the numbers we talking about? we want these small miners to gain 25% of the network hashrate? great then we need 210,000,000 of them.

I know what you are saying is "in theory", I am just saying it's not even possible in theory.

you could achieve something decentralized like a like a torrent distribution network where everyone even with 10gb and 10mbit is contributing to the network.

Torrent is built on a different structure, anything above 0 data and bandwidth will benefit the network, even if it's a 1kb txtfile and an old dial up connection you would actually contribute to the overall streght and availability of the P2P network, in BTC, it's different, if you are not finding blocks -- you are just burning electcity for no good reason.

And as the hashrate grows and so does the need of electricity, how many countries out there could host it?

I believe the opposite will be true, the more times goes buy, the less eleciticty BTC would use, of course, we are not there yet, but I think of it as climbing a mountain, we would get to the top of it and then start going down, the only reason why the market still allowes electcity growth is because we have not hit the effiecny wall and the rewards are still huge, eventually, when there is a little room left for hardware improvement, and the block rewards are next to nothing, there bitcoin ecosystem won't have a room large for everyone.

Obivouslly, given that we don't know which gears are running at the moment, we could have already been in that stage, talk for an instance the S9 era where hashare was in the 60EH, it has grown by a factor of 10 today, obviously, the power consumption did not, and if we were to assume that it was only S9s then (the most effiecnt at the time) and assume that it's all S21 or M60s today, then effiency has improved by 80%.

60EH back then needed 6,000MW (if all are S9s)
600eH today needs 12,000MW (if all are latest gen miners)

So while we increased the hashrate by 900%, we increased the consumption by only 100%.

Obviously, not all gears today are 20w/th, also not all gears back then were 100w/th, these numbers are somewhat random to explain my point of how in the future we would likely be able to achieve power demand decrease.

Russia and I really don't think that having that country host  an important share of the hashrate would be that much of a great idea

Is that a politically biased opinion or there exists a valid technical reason behind it?

But regardless of what you think about said country, I doubt you think they are worse than China, do you? the Chinese had control over bitcoin mining, and nothing happened, personally, I don't believe "evil" would exist in the mining business regardless of the government views of BTC, I highly believe that rich people are stronger than all governments, if there exists a country where a few rich people own 20-30% of the total hashrate, it means they are strong enough for any government to screw them over, I know what some legacy media might tell you about countries that are ruled by a single man -- they say the same thing about my country, but I know such things don't exists, power resides where money resides, the government (or otherwise a single handed man) is only strong enough to control the average citizens, the rich -- will always operate as they please and will always do what's best for their investment.

So ya, we could say x country is bad and y country would be a better option for BTC mining, to me, I don't care what the second country is, even if it was in a different planet, it's certainly better than all of it being in one place, but again, as we know, people from different countries/planets can always come together, sit behind closed doors and do whatever the fudge they want to do.

Build a 1 hashboard s19 in a custom case.  cost is under 500 bucks

does 30th at 900 watts.

This works for many people in cold areas.

I Think I will be making a demo unit  will post it up soon.

But as much as I know thousands could use this in the USA I doubt I will be able to more than a few hundred of them.

[stompix]

Is that a politically biased opinion or there exists a valid technical reason behind it?

Lol, the hundreds of thousands of Russian coming to Europe with the same stories of corruption and government taking their chunk of business, with agencies demanding they would hire people that would no nothing just stay there and monitor how much you have to pay, with the whole weaponizing of every trade, with the takeovers of business on a wimp, you think this is political bias? 
My company traded with russian companies before, you know how funny is to give full paid vacations to people that don't work for you? 

But regardless of what you think about said country, I doubt you think they are worse than China, do you? the Chinese had control over bitcoin mining, and nothing happened, personally, I don't believe "evil" would exist in the mining business regardless of the government views of BTC,

Well nothing happened at all for years, and suddenly they wanted miners out! Could anyone have done something against it? No!
The fact that something worse didn't happen doesn't mean it won't , it simply means they didn't felt like doing it!

I highly believe that rich people are stronger than all governments, if there exists a country where a few rich people own 20-30% of the total hashrate, it means they are strong enough for any government to screw them over, I know what some legacy media might tell you about countries that are ruled by a single man -- they say the same thing about my country, but I know such things don't exists, power resides where money resides, the government (or otherwise a single handed man) is only strong enough to control the average citizens, the rich -- will always operate as they please and will always do what's best for their investment.

Jack Ma probably would have a story to tell about that! So would the dozens of Russians businessmen that jumped from a window, or the Lukoil chiefs that decided to kill themselves with their families the same day in different countries on vacation!  I don't know exactly what county you talk about but there are levels of corruption and distribution of power, your country could jail people for 25 years for insulting the head of state and still not be as rotten as those.  It was the same here under the previous CP, eveything was fine till it wasn't and declared a threat and my relatives ended up for years in jail because of suspicious "roots"! They too might have a story to tell you but unfortunately I can't arrange even a spiritual session, cause even now we don't know where they got buried! If...
It's the same thing that happens to some in quite freely run states, one tiny abuse and they think they are as worse as NK, but they've never actually experienced even 1% of what's there! Trust me, you've never actually experienced true dictatorship, or at least you haven't been on the wrong side at the wrong time!

Back to the miner, don't get me depressed on Saturday night!

Ya but that is a lot of "ifs" which we all know isn't going to happen, there are not enough people (millions) who can afford and care about "mining decentralization" without caring about the cost/profit, also given that we talking $100-200 per th, it just not going to happen, even if you had millions, speaking of the device in the this very topic, it does 0.7th per device?

Wait, let's not go for zeus price but for prepackaged chips!
So not $40 but 5000/324 so that would be $15.
Again, I don't know shit about the rest and the prices for it , Skot was still active lately here so he might throw around a number and I think we could drop the above quota to 4 or 5 if we talk massive manufacturing and not 100 orders!

so you are going to need 1.4M of them to obtain 1EH which is not even 0.2% of the total hashrate, so what are the numbers we talking about? we want these small miners to gain 25% of the network hashrate? great then we need 210,000,000 of them.

1) I was aiming at things a bit higher, that would be 40-50W so 4-5 times higher and shaving 20% at least after the halving.
2) How long would the reward be able to keep that massive hashrate online? The price won't double forever, at some point all the extra things (land,rent,wages) will take their toll
I won't say it's doable now or even remotely in the next few years but I would rule out the word impossible!

in BTC, it's different, if you are not finding blocks -- you are just burning electcity for no good reason.

I don't understand why you're fixed on solo mining and why those couldn't be part of a p2pool?
But ok, let's drop the torrent and switch to fold@home example!

I believe the opposite will be true, the more times goes buy, the less eleciticty BTC would use, of course, we are not there yet, but I think of it as climbing a mountain, we would get to the top of it and then start going down, the only reason why the market still allowes electcity growth is because we have not hit the effiecny wall and the rewards are still huge, eventually, when there is a little room left for hardware improvement, and the block rewards are next to nothing, there bitcoin ecosystem won't have a room large for everyone.

Which brings us to point 2 in my above reply! 
I think that after another halving the ratio of what % of the hashrate you can buy for $100 will change dramatically, of course counting inflation.

But as much as I know thousands could use this in the USA I doubt I will be able to more than a few hundred of them.

The moment a heater cost $50 at walmart and the miner heater cost $500 and recoups only $10 a month probably not even 1% would.

[mikeywith]

Stompy I have so many things to reply to your comment but then we are going to turn this into a political discussion and you probably won't like whatever I have to say. so ya i will stop there and go back to "mining"

Wait, let's not go for zeus price but for prepackaged chips!
So not $40 but 5000/324 so that would be $15.
Again, I don't know shit about the rest and the prices for it , Skot was still active lately here so he might throw around a number and I think we could drop the above quota to 4 or 5 if we talk massive manufacturing and not 100 orders!

If you are going to buy miners and extract the ships, then that can't be a mass production of any sort, you are also going to pay a lot more on labor work to get those chips removed and cleaned to be ready for a new installment, but i understand what you are talking about.

The thing is, you are going to need a controller, some cooling, and you are pretty limited by "how many chips" you can put in a single board to keep it as quiet and easy to run.

So it may not be 150$/th but surely not even close to large miners. It would be very difficult to make these small miners for anything less than 5 times the cost.

I don't understand why you're fixed on solo mining and why those couldn't be part of a p2pool?

It's because I think nobody is going to spend 150$ or whatever it is just to get $2 a month, and also the fact that all of them combined would never reach a hashrate high enough for large miners to give a fudge about them.

Obviously, at some point in the future, things could change, but not to a degree where group of small individuals beat large corps in a business, can you name any other business where such thing actually happens? i can't think of any.

It would be great for home miners to add a little heat to their rooms, and get a shot to win big, but I don't think the impact on the whole network will even be noticeable.

I think what Avalon is doing has a greater chance of "success" or something like what phill mentioned above, something in the 1kw range and 30-40th, but these little experimental cute toys are just.. fudge without chocolate. 

[stompix]

Obviously, at some point in the future, things could change, but not to a degree where group of small individuals beat large corps in a business, can you name any other business where such thing actually happens? i can't think of any.

That's the thing!
Those shouldn't be run as a business!

It's the same argument I had when people were talking how expensive it would be to run a node with blocks at 4-8MB, you should not underestimate what power could thousands and hundred of hobbies have if given the same tools, there were trackers that had with their peers amounts of data that would make datacenters scared of even thinking of hosting,  Folding@home beat all the supercomputers to the exaflop marks! Would any of those be able to come even by renting his space or CPU power, of course no, but when they are doing it for fun and not looking at the cost they have an advantage!
Would I be able to beat a huge company by starting a transport company with two cars I buy on a 11% interest mortgage? No! Can I beat the crap out of every train/bus shuttle/uber driver by offering a lift to my already settled destination, the way blablacar does? Yeah!

If Bitcoin crashes tomorrow to 20k (god forbid) how many of those business would go under?
As for the guy running his 50W fancy lights toaster miner, do you think he will care he makes 2$ a month instead of $5?  

Again, as a business those things are a money pit! But if you don't care about that, the whole thing changes!

It's because I think nobody is going to spend 150$ or whatever it is just to get $2 a month, and also the fact that all of them combined would never reach a hashrate high enough for large miners to give a fudge about them.

"Nobody" is way to harsh, not all of them currently being able to make a dent yes, indeed perfectly true, as for the future it's matter of money!
When farms can't get a ROI in 60 months on 4 cents per kwh then it's time for the dent in the marketshare!

I think what Avalon is doing has a greater chance of "success" or something like what phill mentioned above, something in the 1kw range and 30-40th, but these little experimental cute toys are just.. fudge without chocolate.  

1kw might be cut the numbers of hobbyist needed, but it also cut the number of user.
You need something that you don't care it's on or off, for your average Joe that would be something under 100W, well ..depeding on the region.

[mikeywith]

That's the thing!
Those shouldn't be run as a business!

But it is a business to many people, you may treat it however you like but that won't change the fact that large corps treat it as a business because profit is involved, it's good that you brought up the full node part because it's likely that the majority of nodes are run by individuals, even the same people who run many of them do it for a good reason (better connectivity), but if it had it any direct benefit "money" then large corps would own most of the running nodes -- regardless.

So until we get to the point where the entire market of Bitcoin mining is incredibly small to the point where no large corps would be interested in it -- home miners will not be able to compete, but when will that time come? I don't think I'll live long enough to see it.

"Nobody" is way to harsh, not all of them currently being able to make a dent yes, indeed perfectly true, as for the future it's matter of money!
When farms can't get a ROI in 60 months on 4 cents per kwh then it's time for the dent in the marketshare!

"Nobody" was likely the outcome of my limited English vocabulary. Arabic is my mother tongue, so I believe my restricted English proficiency should be valued by you.

But yeah, I don't mean literally nobody, because then all it takes is you running that miner to prove me wrong, . I just meant, not too many people.

You need something that you don't care it's on or off, for your average Joe that would be something under 100W, well ..depeding on the region.

Ya, you are right, something like an old 100w light bulb that won't break the bank if you leave it on 24/7, but then again, 1000w for something that would actually make you some $$ would make more sense to some people, every house is different, so getting the best number is not going to be easy, anyone planning on doing so would need to make 100w, 250w, 500w, 750w and 1000w then see which one is selling the most, if i was a home miner I would get the 1000w because I can safely run that without going broke.