Quote from: patrolman on June 01, 2015, 10:15:39 PM

Quote from: OrphanedGland on June 01, 2015, 01:15:38 PM

Quote from: Sub-Ether on June 01, 2015, 11:09:07 AM

Quote from: johnambush on June 01, 2015, 09:57:48 AM

Quote from: noobtrader on June 01, 2015, 09:24:00 AM

i now mining dash with gpu, and i want to know if old x11 miner can be used to mine dash ?

edit : is x11 asic miner exist ?

Of course...I'm also mining.

I think x11 asic miner is still no present. But maybe I'm wrong.

A fair point to bring up occasionally but..
the question is rather, is it economically viable to design and build an asic machine for X11?

I designed some very tiny micro controllers and FPGA's in University for projects and we built a fast chip that could multiple two 4 bit numbers together.
You might think that its easy and can be done in 5 minutes, but it took days of involved boolean algebra, testing, errors, checking the faulty logic for latency overlap problems etc. Basically you can make any (speedy) function only from Nand gates and flip flops but you will need very many of them and they are time consuming to setup.

So how long do you think it would take to design a logic circuit to do 11 hashing algorithms from scratch ?!
The Sha256 maths on paper takes 16 minutes, and this is way more complicated.
https://youtu.be/y3dqhixzGVo

The months of development is not the killer issue though,
This is going to require a lot of logic in series and some of the hash functions will be quicker than others so you will have most of the circuit not doing anything for most of the time hence it will be costly to have relatively unused circuitry(more than common GPU's most likely)
You could optimize it of course in order to use some of the redundant circuit but this will bring in a new round of testing and errors (using the same part of an algorithm twice is risky as they may over lap-must be thoroughly tested)

So months down the line and 50k-100k later (investors getting nervous) drawing closer to an official release, you find that AMD have just released a new graphics card that is twice as quick as the last and Wolf0 has optimized the .bin again (TIA), thus quadrupling the hash rate so your promised pay back time has now gone up by 4, not such a hot seller on ebay any more.

And even if you get past all that, it has already been said that if there was an X11 asic machine developed, a fork could be taken immediately and now we have x12 instead and all that hard expensive work at the R&D labs for months was a complete waste of time (for this X11 coin anyway). Would any company take such a risk, with that in mind ? (feathercoin done this already)

So do you take the first option, or do you go out and simply buy a few cheap Nvidia 750 ti's and wait for the next upgrades?

https://dashtalk.org/threads/x11-mining-optimisation-project.2584/page-6

I have personally designed an X11 ASIC - well the RTL only, which can be used for either FPGA or ASIC implementation. It includes a mechanism to account for variations in performance between each hash type. Unfortunately the economics are just not there yet to get this turned into a layout.

If you were to submit a proposal for a Masternode vote (when it becomes available) to release an open-source FPGA or ASIC miner it might be possible to gain the support necessary to achieve funding. I don't have the figures in front of me at the moment but I imagine it would require quite a rally in order to be economically viable.

We are potentially open to this, however investors have spent a fair amount to get where we are and its unlikely the development fund payments will be able to cover it. The community would most likely vote for funding to be spent on protocol improvements rather than mining infrastructure.

OrphanedGland

Posts

Merits