I sorry if I am posting this in the wrong place.
I would really appreciate it if the foundation could acknowledge this post and if it will be considered with their ASIC resistance investigation.
Technical and political aspects of ASIC mining aside. I am trying to address the decentralisation and the availability problem that asics may introduce.
To keep this post fairly short I have cut a lot of detail out. I feel it still keeps the intent I was going for though. feel free to ask for clarification.
The greatest non technical obstacle I can see to the acceptance of asic mining hardware is the distribution of the machines. This leads to all sorts of real and perceived issues.
Bitcoin faced a similar thing back in 2013 when asics were generally only available from a handful of vendors, but lots of vapoware / working prototypes with no real way to mass production.
At that time a number of companies tried or did sell the reels of the asics. This seemed to workout okay for everyone except the little guy. You couldn’t buy 1 chip. only reels at around 30k each (3,000 chips per reel), let alone find a board to put 1 chip on. (this was not an issue with FPGA’s)
Here is a link to a bitcointalk auction for two reels of bitfury ASICS to give context to what I am proposing.
I havent done the conversion maths though. so I cannot really say what this would be in equivalent hash/watt/cost today. It is the idea that is important.
Zcash co&fo is in a much stronger position in regards to hardware than bitcoin was. They are a fully fledged company and non profit that has the founders reward incentive. Apart from that the market is now a lot more mature and the idea of bespoke hardware, while contentious, is accepted.
The foundation has the potential to make ASIC equihash miners commodity hardware. Because of the way the foundation is designed they can make business decisions that would be ‘company confidential’ in a normal business. However the foundation can do the work, then release it all for free and still see a major net benefit to the community, coin and foundation.
I would like to foundation to consider the following business model? (calling it a business model is a stretch but I have run out of words) Please dont beat me up on these numbers they are just for illustrative purposes.
it would work something like this:
- the foundation pays for reference board designs to be made (via tender) [as example of network impact they would be like buying the equivalent to 1x1070 (5chips), 5x1070(25 chips), 5x1080ti(40 chips) are to the network today in terms of capital outlay cost and coins per watt]
- The foundation would release these designs to the public in an opensource manner. They could also partner with companies like Olimex so they can sell individual boards at cost price. note: these are boards only.
These board designs can also be supplied to ASIC designers for compatibility reasons, and to allow more advanced companies to produce a more optimised product.
- The foundation works out deals to buy asic chip reels from companies like bitfury/bitmain/etc - they sell these in small quantities that match thier open source miners. so I can buy 1 or 100… for example. however it would probably work out cheaper for major players to buy off the bespoke board designers that might put 200 chips on a board.
- The foundation works on supply deals for the entire BOM for each type of board and offers digikey afflation or similar.
-The foundation works on supply chains for other potential limited hardware (ddr4/controllers/fpgas/arm cores/etc)
The foundation also offers a reference design fully assembled version at cost using a supplier like olimex or pcbtrain, etc.
This only really works because the foundation is a not for profit. they would also have to sell individual chips, etc.
This is very similar to how nvidia and amd work with their board partners, it is just they do not opensource the reference design. It is also worth noting that the internals of the asic can remain proprietary, like a normal gpu chip is.
does the foundation see this as a possibility? and if any of this doesnt make sense please just ask.