sigh
OP claims he has money printing machine. OP will sell it to you. Why sell a money printing machine when you can print money ?
~ LOL ~
Ya, ya, ya, I'm well aware of the things I've said... (Assuming you're saying that with the LOL because I've said exactly that myself
)
-- The greed inside...
Because the only way I can get my own money printing machines is to buy enough of them from Xilinx. I can't do that alone. Collective bargaining. Right now we are one cog in a multi-cog collective bargaining operation going on with Xilinx. All of these cogs are competitors. We have all been doing the same thing in secret for some time. But, we all also have the same problem, Xilinx is slow and difficult to deal with. None of these cogs (myself included) would be able to achieve a successful PO on their own. The number of chips we would receive for sales (in batch 1) is less than 1/2 (about 40%) the total amount. Our total order volume right now is approaching $50M USD. Future batches would mostly be made of community orders (80-90%). But there are a few large players driving this and I am not one
-- I'm closer to the community gpu miners than I am these guys with their own private multi-MW hydro electric dams and $20M operations. So, ya, I could let them do their thing -- they'd pay higher pricing... But would
EVENTUALLY take all the profit out of it for my small-ish volumes, they'd take the gpu profits, they'd cause numerous coin forks as a result of their secret hashrates (which could follow and sustain through a fork -- causing various issues in the community). The only way I can see to continue for myself and compete is to open things to the community.
The only way anyone else in the community will buy the money printing machines is if they can print money with it. This means I have to release firmwares and have a method for people from the community who develop better designs than our own to release theirs. This also benefits us as if their design is better than ours, even with the devfee, why wouldn't we want to run it? Our margin on these is ridiculously small almost too small.
I truely believe that the community engineers (people you've not even seen talk in either fpga thread yet) will release bitstreams that have better performance than anything myself or whitefire could release. It's in my interest to provide them a platform to do their magic and provide them with a method to benefit from that.
If you want to invest $30M-$50M in me (at a $200M valuation), let me know, I'll go back to doing it from the shadows!
-- But as I said, I'm pretty sure staying in the shadows would end up decreasing profitability (due to issues, forks, community backlash, etc).
-- The egalitarian inside...
Right now, there are 130, 90, 65, 45, etc nm asics that no one knows about mining various cryptocurrencies. These asics were created by individuals in secret at low cost with efficiencies far beyond GPUs. They do limited runs to reduce their risk using MPW projects, etc. These people, if they wanted to, could effectively 51% attack these currencies and may have -- there has been an increase in the number of 51% attacks on various coins. There has been an increase in the number of double spends being attempted (successfully and unsuccessfully) on exchanges. We need general purpose devices with efficiencies greater than GPUs. Devices that have power / performance and cost / performance ratios closer to ASICs than to GPUs. Having these devices on the PoW networks would secure these networks against 130, 90, 65 etc nm asics. Even 45nm asics may only be slightly better performing than the FPGA devices. This pushes the cost of secret asics down to 28nm which is still millions in NRE and increases investor risk at the same time. A 28nm asic would have a higher mask failure rate than a 130nm asic. The lower you go the smaller the spacings and the greater probability an error will creep into the design.
I'd eventually like to boot every GPU off of every PoW network and to have all the PoW networks without a 28nm asic or better to be secured by a decentralized network of highly efficient FPGAs. The miners don't even need to replace their rigs. They can continue using their gpus for now and swap them (the gpu cards) out over time slowly selling them off on ebay so as not to overload the secondary markets. There's also a GPU co-processor solution that GPUHoarder may be releasing to market. It uses a very cheap and low power FPGA to do some co-processing work for the GPUs. This device can add a little life to your GPU clusters increasing their hashrates and reducing power consumption.
