Topic: Why the Bitcoin community HAS to save the network from HashFast, CoinTerra, BFL. - page 2. (Read 9870 times)

Anyway you can put me in the hat as the first Group Buy Coordinator willing to throw my hat to be a treasurer in if you guys are serious and if this project goes anywhere.

I run a small miner co-op that currently has 18 members with 2 of them knowing the exact research lab and location that I work at. I also vet and coach new Group Buy Coordinators and Bitcoin Miner Hosts for free in the hopes of doing future business with them. I've also been vetted by John K. & he's seen my LinkedIn profile with multiple dozen Hawaii/US Aerospace/IT/R&D professional endorsements and a 162 perfect feedback rating on ebay as my wife's co-seller/buyer of IT hardware. I can also perform escrowed BTC transactions as a BTCrow Escrow Affiliate. I advocate for both buyers and sellers and, in any arbitration, will side with the one having the strongest evidence and least scammy behavior.

My collateral in any financial transaction I do here is: my job/freedom/ability to work on low TRL projects and get paid to do it. Anyone willing to handle someone else's BTC/cash/PP/ID/miner payouts on these forum needs to put up those same type of guarantees to do business around here IMO. And you should use escrow from a trusted provider whenever possible.

OurAsic: I like what you're trying to do here with the openness you're trying to establish as open source software/hardware for 28nm ASIC chips.

I do have issues with the way you worded the release as it's combative and unnecessarily somewhat libelous by stereotyping every manufacturer you listed as evil and greedy when the jury is still out on 2 of the 3.   If I was editing your PR I'd caution that it's been proven that you attract more flies with honey than vinegar. Even if you *are* angry and bitter from being scammed: you shouldn't bring your at-home issues to work is what I was always told.

People are people. While business and politics seem to attract an abnormally large percentage of that "2-3%" that are mentally damaged or impaired such as sociopaths and psychopaths, most folks are decent people just trying to get by as their country's currency continues to inflate (some at ruinous levels), and are people that follow the laws and social mores of their culture, for the most part, eheh.

Perhaps if you had grown up in a hierarchical culture instead of an egalitarian one (relevant thread: https://bitcointalksearch.org/topic/m.3154478 ) you may have found a way to still gain customers by calling attention to current product availability without resorting to negative (and possibly libelous) comments about any particular manufacturer.

BTW, as far as the nature of Greed: I don't want to say "greed is good" like Gordon Gekko but that same greed helped bring our ancestors out of trees and helped them to explore new vistas with all the concomitant unforeseen changes that would be unleashed as we went forward as a species. That same "greed" makes you get up in the morning and go to your job so you can continue to not live in a cardboard box.

Risk taking is good in a species. Extraordinary risk as a species overall, is probably ruinous, but risk taking is what helps makes us human and helps us to feel alive!

will you, maybe make also a 12nm one too (time permitting of course) 

Agreed. It is now the hoard of ASIC's.

I understand your point. In generally I don't have anything against a free market. In this specific case, the mining equipment is securing the bitcoin network, so I'm interested in a decentralized mining infrastructure. At the moment the owners of ASIC designs are able to bring up hashing power with cost of manufacturing, and all others have to pay a much higher price. This potentially leads to mining centralism, which is not a good thing.

To stay in your example with the 28nm/400GH chips for $50 in the hand of the bitcoin community. With such a available design it would be much harder than at the moment to make money by taking the risk of a new design. I think even the best ASIC experts will think twice before taking such a risk. In my opinion bitfury's 55nm design has already squeezed out a lot of the 55nm potential. I don't expect that someone is able to improve the numbers by let's say factor 2 with the same technology. On the other hand a step into a newer technology, like 28nm, is not automatically bringing so much more benefit. The problem is the up-going heat per area, which leads to complex, expensive cooling requirements. I'm interested to see how all of the 28nm announcements are solving this problem.
Anyhow, you can see already now that the cooling efforts are increasing (water cooling etc.) and the efficiency is not getting soo much better. About 0.5W/GH is also reachable with the bitfury design. I would be more convinced about the step to 28nm, if I would see <<0.2W/GH, which doesn't seem to be the case at the moment.  

Let's say you open source a 28nm Asic that does say 400ghz per chip and costs $50 and you sell them at cost. Someone, somewhere will make a better chip if it is commercially viable and then you are back to square one.

People are buying from the current ASIC manufacturers out of free will, there is no monopoly or evidence of a price fixing cartel. There are options for all wallets and ability's (ASICMINER USB to BFL Minirig and more on the horizon). If the high price of entry is such an issue these companies would quickly go bankrupt as they would not be able to sell products if people do not want to buy them.

This is exactly the same as people who bitch about Walmart coming to town and killing Main Street, when it is the people themselves shopping at Walmart that kills the independents. Don't like it don't buy, see how long a company survives with no customers.

I don't think it's so important which technology will be used etc.

The important point of OurASIC was, as far as I understood him, that an ASIC design should get into the hand of the bitcoin community.
So different groups can get this ASIC's for cost of manufacture and can build different, competing miner designs for the market.
Because of the extremely high engineering costs of a new ASIC design, this is not really true at the moment. Only big players have their own design and are selling ASIC's for very high prices.

Personally I like bitfury's ASIC design a lot. It's "only" made on 55nm technology, but very well design and can still compete with the current 28nm design announcements. If the community would have a design like that, I would already be happy. Let's say we would get the ASIC incl. packaging in volumes for about 1.5$, which is not unrealistic. The costs for the chips for a 400GH setup would be around $200. Without the need for complex and expensive cooling requirements like water-cooling etc.

Maybe it would make sense to check if the money for an own ASIC design, could be spend in a better way by just buying an existing ASIC design like the one from bitfury and open-source it.

    
 

 

"save the bitcoin network" isnt a valid argument anymore.

thats early btc adopters + asic manufactures propaganda.

^ agreed.  And "save the bitcoin network!" is a hippie mentality.. I am picturing them chaining themselves up to trees, but in this case it's the blockchain.   I doubt these actions would stop the bulldozers.   Mining isn't for the little guy, plain and simple.  

Because it Lets the free market work its magic. If they can't match the price they fold.

How flooding the market with $100 ASICs suppose to save the mining community?

It will accomplish the opposite.  You'll need $25K minimum to invest in h/w to make BTC1+/day.  Same as today.

People with deep pockets will just hoard the chips and small guys will be left out there with few 500GH/s chips that
will make BTC1/year.  All will change is the arithmetic.  The proportions of ROI/profit/diff will stay more less the same.

I think the rest of this mining ASIC cycle will not be done by mining community at large, but a small group companies who
already got enough money from their gen1 products to go completely solo with their next gen chips.  Think Avalon/ASICMiner and Bitfury.

They might still "offer" border line ROI products, just like bitfury and ASICMiner is doing today.

Let the free market work its magic.   Interfering with it is wrong.  

Sure it is painful to watch ghash.io reap 1000s of coins a day, but it does not really matter who mines those coins.
If you don't like the current ASIC offerings, just vote them out with your wallet.  KNC, HF, CT will be mining with your pre-order money, just like BFL and Avalon did.

Your effort will only speed up the process.  
Bitcoin is an example of ruthless (or real) capitalism (not the gov sponsored capitalism we have today).  Let it be, it does not need fixing.

Lol, not much smaller than 28nm atm. The ROI on anything smaller will be terrible, way to bleeding edge.

This seems to be the popular opinion right now.. but if Bitcoin starts tanking, you're all going to blindly jump into mining again right?   

Very interesting and I am interested to support the effort if the details all line up.

My biggest general concern is "are we too late" when it comes to keeping up with folks who are basically at 28nm now, and moving to smaller soon...

fuck them all.

just dont buy them.

if asic manufactures are asking for so much money for their ASICS, leave it to them to secure the network, and just buy and hold bitcoins.

Cant you see what nonsense that is? Mining companies will sell every asic they can produce, the only variable is price and shipping date. Its impossible for them to sell at a lower price and still get the same revenue, thats laughable. Lowering prices will not help a thing for anyone, since these companies are already selling out and lower prices would increase demand which they cant fullfill.


Cointerra started lasted week?  More nonsense.  The company was founded in April and doesnt expect products until the end of the year.
If anything, 6 months for a 28nm design cycle is a very aggressive schedule.

if it were so easy to design an asic, everyone would be doing it.  its not a week to design an asic.  im pretty sure cointerra and hashfast have taken (skilled, experienced asic designers) something like 6 months, certainly closer to 6 months than 1 week

since youve got vmc and others offering a 16 GH chip in 28nm, and knc offering a 100 GH chip in 28nm, and hashfast on 400 GH and cointerra on 500 GH.. then clearly, the design of the chip is the thing that differentiates the performance, and i think the performance these guys are hitting is represenative directly of the skill of their asic designers (and the time they've taken to optimise the power consumption).  thus its not just any old sha design thrown together and it seems like the skill of the engineers does actually make a difference to the design and enables the consequent performance.  and to achieve your desired goal in the 400-600 GH range, that is a very high end, state of the art chip, and will not be achieved just by throwing together some public domain sha engines without any effort or skill required, - if it were true anyone and everyone would be doing it.  in order to reach that level of performance you will need to do serious engineering to lower the power consumption and improve efficiency or you will butt up to the thermal limit of the chip without hitting the performance goal

May I ask how do you plan to effectively control the rate of difficulty climb?

Explain it to me, because I don't understand. Even if you are extra-ordinarily cheap by the time you release your ASIC, the prices from other vendors will probably already be matching yours.



For example, I created a CoinTerra chart a few days ago as to where their prices should be as difficulty climbs. If you look carefully by the time you ship, you will have to compete with them at their price point.

I don't quite understand how your company plans to lower ASIC prices (assuming this is the plan) to stop the crazy rise of difficulty?

http://s9.postimg.org/hcigp57fz/Terra_Miner_4_2_TH.png

if you think you can sell a chip that does 400-600 GH in 28nm, at $200, and somehow make a decent profit, youre wrong.  these chips, running at, say 0.6 watts/GH, are generating the top end of heat that any chip is allowed to generate (250-300 watts!) thus these chips and their packages, and their high performance cooling system would cost more than $200 each.

Then youve ignored the cost of the nre?  that needs to come out of the cost of each chip, ie: amortised over however many chips you plan to sell.

if youre going to sell 10,000 chips (which seems like a lot), then divide the $3m+ nre cost, over the 10,000 chips (ie: add $300 on top of the cost of each chip, before you can sell it at zero margin.  lets say it costs $250-300 each, for the die, the package, and the cooling system).  plus $300 contribution for the nre cost, and youre already at $600 minimum price that you could sell each chip - at zero profit margin !

-- Jez

Very interesting post. It coincides with a lot of my ideas about the future of bitcoin mining.
I would love to work in any kind of way on a fully open source ASIC project.