Build Your Own ASIC Miner


Hey Guys,
It‘s very nice the idea, but it’s not like this… I’m IT-Specialist and Miner, to produce a chip, you just need a team, but not like old school as apple ideas from old Xerox and inside the Garage, ASIC Chips are not complicated but are small, (nano) small, what it’s pratically impossible to reprogram, because you don’t have the IP-Layout, you can not program or reprogram an ASIC so far as you think, it will be nice, but for that, there is just two ways, one is to crack the code, and there is no hacker in the world that can do this as days or months… or years, maybe… and maybe is more easy to hack the company and have acess to the IP-Layout, maybe I’m wrong, but you need acess to a manufacturer Nano chips, and if you want this you need something like 40 People (chinese) and expensive machines to produce ASIC, but you will need fly to China, to find the best cheapest expensive prices, about 5 zero or 6 zero dollas… yes, than you will discover that you are working with something 28 nm 16 nm, and it’s already old, because they are producing 10 nm chips… yes… it’s better to forget this…

But if you are really wanto to produce some boards with custom design, you need to contact Bitfury, these guys are the best and most reliable than Innosilicon, who destroyed the pioneer of Miner Production from GPU to ASIC, the Bitmine AG (just google the name)…

Man, if you want to recycle some old miners, I just suggest you to think as real programmer and real chess player, find another way to do the check mate, as an example find a big place and produce free energy, as a solar energy… and them you can use old miners… and relocate them together… use something like this…

Or create your own cryptocurrency… but today is better to do a token… or something like this… yes it is not easy… sorry about that…


And @FloridaWater believe me, I have total respect for your knowledge, but the decentralization or centralization comes not from a place but from who operates, I Could you give more details here, but sometimes we can not talk so free as we want… there is more than you think… and you will understand, only if you travel to Chingling and there, just there, you will see, and see with your eyes… and understand what I’m talking… and everything what you are proposing, you will discover that, maybe, your enemy, is your friend… maybe your best friend, or professor, or your chief… who cares about centralisation and decentralisation now? This is the question. Who is caring?


I’ve already got the chip designs, and have the money to put them into production at a foundry. The main issue is all of the regulations, and legal liabilities you face selling them as an American company.

I even have a prototype that is using high end fpgas. I’ve got the skillset, and knowledge. I have an ex Intel motherboard designer at my disposal too. Competing with the Chinese is where my work breaks down.


What about partnering with someone who already has compiance and cutting them in on it? might be cheaper then getting compiance yourself


I want to open source it. It’s not competing with the Chinese that bothers me, as much as having to rely on them.

Custom ASICs can’t really be open sourced, because you can’t order the amount to build one machine. I’ve considered kits, but Butterfly Labs did that. I think most of the failures we’re their fault for not planning well. They’re still suffering the costs of it.

If I could talk a chip manufacturer into producing them for sale to hobbyists that would be ideal. I don’t think they’d let me realease the IP if I could make that happen.

We’re losing all of the expertise in hardware design because of cost. We have coders everywhere, but not many people that could design a chip. There isn’t even a complete version of an open source GPU that I know of. Microchip companies have been killing it in the stock exchanges though.


Sounds interesting. Maybe you would like to make your organization a non profit research entity so as to avoid some regulations and legal liabilities. Also, instead of incorporating as an American company, there are other countries with interesting tax laws.


Opensource GPU?



what about building them to run in your own DC in the US? is it the same liabilities as if you were selling them or no?


I’ve been thinking a lot about a non-profit. Lack of free trade hurts any other country other than China. That’s another issue.


I don’t think either one ever made it past fpgas, or were full fledged GPUs. Just parallel processors. I might be wrong.


A rogue nation has crossed my mind.


i am sorry for not being online for so long
i am not a software guy ( my coding skills are less than 1 out of 100 )
however i have found a source of bm1382 chips along with the data sheet and the schematic diagram and this is all i need to get a functioning board bcuz they are not that hard to work on with
but what i need is a guy that can link these boards to a mining pool and that’s it


What type of interface are you using for the board? If it’s parallel, a Raspberry Pi should do it. It depends on how you plan on setting up the clock registers, and how many chips you are planning on using. Just buying an S3, and taking the boards out of it would probably be the simplest way.


Thanks for all of the good input here guys, I like how this discussion has slowly started to develop and I like where we are heading.

I have not been doing a whole lot more of technical research on how to build ASIC miners since I started this forum. However, I’m glad that everyone is sharing useful knowledge here. We have a decent think tank going here.

From a business perspective, I think customer service is super important, as product design. Like Peter says, if you don’t have a well designed product, you don’t have a business. Well, Bitmain has a supposively great product, perhaps the best in the industry. However, Bitmain’s customer service is the worst I have ever experienced. Long story short my D3 ASIC miners were shipped to Alaska, didn’t clear customs, then were shipped back to China. I am still trying to get Bitmain to resend me my miners. I have made numerous international purchases and never had any issues with customs besides some trademark issue that was quickly resolved. The fact that Bitmain didn’t provide the correct documentation for customs has lead to this unfortunate incident.

So, not to slander Bitmain, but to say that we could totally copy their product like so many Chinese manufacturers do and then just add our stellar U.S. customer service to our product. We should not have to rely on China for ASIC miners.

We have the intel, we have the skillset, and knowledge. We can compete.


Their customer service gets better depending on the size of the order. My main issue is that they own the market, and they aren’t very friendly for home use. Tearing down Antminers and making smaller home versions from the hashing boards would be a lot more cost effective than actually producing new ones. That’s not what I’m looking to do though. I have considered it.


Okay, @Jax2015 was advertising to sell 10 of his S1 antminers. These could be good to experiment with.

I am still waiting on my first two ASIC miners to arrive. I kind of wanted to order an S9 but decided against it when I experienced the horrible customer service with my first order.


They’re simple, but not really efficient enough to run anymore. You can get the basic design cues from them. Building a main board is the hardest part, aside from having chips produced. I don’t think it’s really worth doing for the old chips. Modifying the old boards for better performance, and efficiency can be done.


Okay, so maybe just one S1 miner for design cues. I might buy one.

Have you checked out any of the boards and chips on


I have, but it’s hard to find a legit supplier on there. I ordered a reel of tegra chips earlier this week for some experiments. I’m waiting to see what actually shows up.