Build Your Own ASIC Miner


#61

actually i would say you are on to something , example DASH, the price of dash is going up and up because it is so unprofitable to mine with the d3 miners , so people are holding more and more until the price goes up.


#62

Btc Price will moon this sunday anyway after the 3 “super forks” and the wall street debut in chicago. will moon to 25 k or more, theres no need to be a genious to understaind that as more publicly BTC goes more investments humps in and market cap grows. But your theory could be put on practice on any btc price. Nice thinking there.


#63

I came across this page via google search, was trying to do the same thing. I’m in the process of reverse engineering (at a high level) the AntiMiner. So far it looks like a box with several asic cards, fan cooled, run by a beagleboard, and a single i/o controller board. + a decent power supply. [I have 50+ server power supplies looking for a home or to be put to good use]. There is no reason this can’t be done with a pi and asic hardware.

I/O Board: https://bitcoinware.net/collections/antminer-asic-mining/products/io-board-for-antminer-s7-with-3-connectors-new

Single Board Computer: https://bitcoinware.net/collections/antminer-asic-mining/products/beaglebone-bb-board-for-antminer

Asic Card: search alibaba for mining cards (i can’t include any more links [2 link max per post])
A final word on designing your own asic card: check out nandland dot-com they have a great article on using your own fpga to mine


#64

Hi colleagues!
We are an Austrian/Europe company and are ready to start big mining business. We are cooperating with hydro power plants at low energy cost.
As we want to size our farms with about 15 PH/s and we are ready to start with 10 farms, we are little confused about the BITMAIN and BITFUR policy and pricing. It is nearly impossible to buy that miners at reasonable prices.
So we thought about “designing” and producing our own miners. This would take months and we will lose a lot of opportunities.
Is here anyone ready for prototyping such a miner?

Please tell me, how I can approach a test with any of that stuff.

We want to prequalify our equipment and want to see how our immersion cooling system is working. We will cool the asics with a special liquid and so they will run at max. 60°C

Hope to read some interesting issues and proposals.

Regards, Chris


#65

I like how y’all are networking, establishing relationships, and choosing to look at the simpler parts of design first, with faith that the more complex details will be discovered along the way, and solutions soon to follow suite. At the same time, we are aware that this is not such a simple undertaking, but with time, effort, the necessary framework, teamwork, and resources it can be accomplished. Cheers to those who are already establishing relationships with large renewable energy providers, very noble. I like the possible can do attitude and idea that this can be accomplished with a pi and ASIC hardware as well. I will order a raspberry Pi this week as I continue to negotiate with Alibaba vendors for electronics and Bitmain customer service to have my miners reshipped ><


#66

I propose that the Austrian/Euro company negotiate with your hydropower plants to use their renewable energy to mine with ASICS, and that you provide the ASICS miners. This is the idea, right? So, if you have investment capital, I would start your farms by buying new and lightly used ASIC miners on websites such as amazon.com and ebay.com. For example, I bought my D3 ASIC miners for $1300 each, but some people are selling their D3 ASIC miners new in the box for $800 because it turns out that they do not have the proper infrastructure to set up the miners, they don’t believe the miners will be profitable to them, among a variety of other reasons that may benefit you under different circumstances.

Yes, then you can also design and produce your own miners, improving the flaws of the previous designs of course.

We are preparing to prototype such miners here in this discussion, collectively on good will, for science, for business, for the profit, some not for profit, for the pub, but also decentralized :slight_smile:


#67

Thanks for sharing these ideas and resources. I hope that you can put your server power supplies to good use in good time. I’m glad that you have joined us for discussion here at the pub and I hope that you will continue to visit.


#68

Bitcoin using blockchain was supposed to be fair for everyone. This monopoly isn’t !
I go all the way with you Nathan and Charl , make it open source let everyove have a taste of the Pie.
I believe in you guys.


#69

Regulations are not that rigid in India, drop me an email on sumit58262@gmail.com, i like your idea and i think we can figure something out. Decent Miners arn’t available and the antminer’s price is a kill and i’d like to help.

I am a law student and a believer in blockchain technology and decentralisation.


#70

I like this idea very much. So I see everyone is taking a very difficult manufacturing edge to this. I prefer simplicity. So instead of building a rig that is custom like the antminer - why not build a PCI board version of it. ASIC SHA256 chips are pre-made. You order them the same as antminer or the others do. Making them work is typically a controller chip of some sort, and memory bank.

To prototype this instead of thinking big, think small. How would we build this to work on a simply PCI-E bread board to a small scale? The key is scale. You can proof a workable model on a pci board, and when it comes time to build en masse, simply scale the number of asics up.

Im my mind what we need to make this work is:

A) an electric engineer: familiar with PCI bus technology
the asic chips in a scalable pattern, and etch them to the controller chip
who can work in an appropriate buffer mechanism
B) a computer scientist who can right the necessary drivers / interfaces for the system to communicate properly with the custom card.

Once you have a small scale card working, simply scale it up by adding chips, or by using later revisions of said chips, and then take an open source version of mining software and adapt it to use the specialized hardware.

I like the PCI idea because now its more like a GPU - a card you put in your system - but instead of processing graphics its just hitting sha256 specific asics for the processing. This makes it easier to mass produce in the long run as well.

Just my two cents - message me if this gets off the ground - I would love to be involved and have a computer science background.


#71

This may be a source of the asic chips.

http://www.heliontech.com/downloads/Helion_PB_-Hash(Fast)_ASIC.pdf#view=Fit


#72

Mechanical engineer here,
Graduated from Nottingham as a degree holder
Currently pursuing masters in UM,
I can help out in designing heat exchangers for the ASIC Miner, lemme know if you need help with that. I’m also interested to know how to build one decent one with less :moneybag:


#73

Hey there, it sounds like you know exactly what it takes to make this happen. Thanks for sharing so many details! I am going to put some job announcements out on craigslist and indeed.com for some crypto start up positions. I honestly don’t have the resources to employ high paid guns like electrical engineer or computer scientist, but hopefully we can open the discussion of developing ASICs as a start up opportunity, and try to get them plugged into crypto for gains if they’re not already involved. :slight_smile:


#74

Very cool that you have successfully graduated as a mechanical engineer and are actively pursuing your masters in UM :slight_smile: Congratulations! I think that this topic could make for an interesting master’s thesis topic. If I am accepted to graduate school for business/cyber security/non profit management, I hope to incorporate this project.


#75

Cool, reel me in bro hahaha.
I have all the equations needed to make the heat exchanger.
I agree, it is and interesting topic for Master’s thesis topic but its
gonna take a lot of time to do it since there are other disciplines
involved, electrical and electronic, not gonna be easy though.

Keep in touch man,
i wanna learn how to build my own asic miner although it is near
impossible.


#76

nathan -

can you shoot me an email - tornadoslims@gmail.com. i have the resources you need in the US and China.


#77

Gents:
I’m just catching up on the conversation here. I am a mechanical engineer with tons of experience in thermodynamics and air flow related acoustics. I will help with the design of the cooling system to make it as energy efficient and quiet as possible.
I can’t offer a ton of help with the chip design or software but I have some friends we can drag into the conversation.
Let’s keep this going!


#78

While we’re at it, we should be thinking about how to get these decentralized miners on a pool that is as fair as possible. I’m new to the game, with an Antminer S9 running for just a few days now. The biggest mining pools are too big (anyone who approaches 50% is a threat to the integrity of whole block chain), and the smaller guys have a hard time providing reliable payouts. With decentralized mining, we will need a way to scale a decentralized pool too.


#79

I am new to Bitcoin mining as well and willing to out all of my knowledge and resources into helping with the creation of an affordable asic miner. I have a great experience in programing and I am good and problem solving.


#80

Sure do. Just takes a bit to get something like this off the ground. Let me know how you make out - I have a computer science background, and can probably code the necessary drivers (linux).

What we need is an EE guy to put the hardware together. I have a neighbor I can speak with who is an EE professor at a local college - perhaps he would like to be involved. Also, the ASIC companies themselves (who want to sell chips) probably have predefined patterns we can use for the etching and configuration. Before hiring anyone I would reach out to them and tell them what your looking to do (source an asic partner, and develop a R&D spec). Because they want to sell chips by the millions, they are usually inclined to provide R&D help based around their product.

I like the idea of simple PCI cards - I have 2 antminer s9’s currently, and all these things are is 3 asic boards (each board has 63 chips on it), that are connected to a very basic ARM based device. I havent gotten as far as to figure out the buses in use, but I would be surprised if it was anything more than USB -> PCIE.

So if we can design cards that can be put into a PC - we can then scale it into a box like an antminer (or something better IMHO).

PM me if you want to talk about getting this off the ground. I am game. I would have 40 s9’s right now, but between the lack of supply and price gouging I am looking for something better. Perhaps building our own is the opportunity.