Bitcoin Forum

HI Guys

 as everybody should know ,i m trying to make my own asic.But costs of NRE is making things difficult.

Recently i ve been working with a design house who came with the following:
 of an asic based on 65 nm and with a hashing speed of 0.7/0.8 gigahash/sec.

With a QFN package the NRE(design+full mask+packaging+design of pcb) will be around 900 000 on wich we ll get around 120 000 chips .

These chips will give us 84 terahash/sec this mean a our cost for the first 84 terahash will be around 10700 usd and a cost per chip of 7.5 usd/chips.



My idea is the following:


People who are ready to buy the first 84 terahash at the cost price of 7.5 usd/chip or 10700 usd/terahash will have the possibility to order twice the hashing power (in chips) 3 usd/chip or 4280 usd/terahash.

The next sales will be at a discounted price for the first comers and kickstarters.


P.s i ll put myself 20/100 of the amount to get this done.

The design is almost ready and we need now to go to tape-out and have the funds ready as soon as possible.


I m thinking to put this on a crowd funding website to get this done properly so any suggestion at this point would be welcome.


I would prefer to have limited participant in order to get some wholesalers  who then we ll distribute their chips as they like but i don t want to limit anyone in this

Please forward your comment/thought here or thru pm

thanks

I'm game, I think this entire ASIC bullshit is just that - BULLSHIT


Avalon takes $8million+  and delivers nothing.. screwing over Western customers...


ASIC's are pretty much a scam unless you in on the design side

Hmm

Noone seems interested


Well Well

crowdsourcing at attractive prices seems like a good idea but everyone knows the proper way to make asic miners
1.) is to design something that works really well and is currently profitable
2 - a.) mine with them until you make back your money.
2 - b.) take pre-orders posting current mining stats as "speculated stats" or "prototype stats"
3.) fulfill the pre-orders and sell them for just under what they would ever in their lifetime make back.

if you upset that balance, then a lot of people will be out of jobs in the mining hardware scene (aka making it an honest trade)

so far that seems like the business plan of the most successful mining hardware vendors. well that and just selling older hardware at prices that will just barely break even before being useless (cost of power > btc output)

I have no proof, just speculation and observation of current scenarios. Only time will tell with all the current pre-order hardware. but it is always very convenient that the hardware that comes out will just barely break even on mining. then the mining companies just say "its an investment into bitcoin"

The only way to make money on btc mining seems to be buy big, mine a little, sell the hardware before its useless to hopefully have coins while breaking even. and the only way to make money on bitcoin (apart from selling hardware) seems to be to buy low / sell high on the exchanges.

Too little too late. $900k for 84 TH/s several months from now best-case scenario? Buy 10 reels of Bitfury chips and you'll have them next month doing 80 TH/s. Far easier and more profitable.

Compare the price,we are ready to tae out so chips will be out in 2 months or so

how much is the cost of 10 reels?

just make your calculation

Is it possible for the design house to go direct to 28nm or smaller, and for the net device price to be $3 per gh/s or less for delivery in 4 months?  That's what you're up against right now.

the cost will be much higher if we go for the 28nm


at the moment this is the best deal we ve got.

Great idea, but too little, too late unfortunately. Those prices just aren't competitive anymore unless the chips were available TODAY.

make your own calculation

in my eyes it still doable and profitable

I've spoken to a few.  While 65nm could work you are probably better off teaming with someone like bitfury and buy a bulk of chips for cheaper.   I mean that you could use a hundred thousand or more and bitfury's tech to have chips packaged at a high volume discount.

As I speak now I know 3 or 4 companies already doing 28nm.

The next generation chips will be 28nm or even 14-20nm.  If you do come up with 65nm you are coming up a little late in the game.  Especially being that there are already multicore ASIC chips out.

Why not decide amount... buy ready chips when available at good discount (bitfury, cointerra, bitmine etc...) and team up with some assembly and find cheap electrity place for hosting ready units and share income based on investments? Payments just before chip buy and assembly etc. just small % for covering managing & hosting etc.

10 reels of bitfury for 30k total chips is ~675k euros as a heads up, unless there is a bulk discount.

How can it compete with Cointerra which is only $3500 for 1 TH/s?

Well

after the mask is paid ,see my post the chips will be very cheap so the first run will be higher then it ll go dramatically down

so there is big chance that we ll be still be profitable

and there is an important difference:people investing here will be partners with us and not customers

comparing to other projects

ok then i ll have to pay him again ?the same amount for having more chips?

here once the mask paid ,i intend to sell chips at costs price (greed is not part of me)

bitfury has an agreement with me he didn t fullfill , i ve posted about it several times if you have read my previous posts

so i m not seriously interested in working with someone who doesnt respect his own comitment

i hope this is clear

I wasnt suggesting we go with them - just putting number our there for everyone else.

ok got you

i m opening the same thread in securities

i hope people will have more access on this

Who would be doing the pcb for this?

details of the chip will be provided so anyone would be able to make a pcb as it suits them

which most people don't have the skills for of course. good luck tho

again difference here is that you ll be part of the project and not customers only and you ll get chip at costs price(after the mask being paid)

there is huge difference between this project and just throwing money to some company to buy their miners

this is a draft of the lay out of the chip

http://s10.postimg.org/oacdjqcft/Layout_Photo_draft.png

some kind of middle and western or new age art work, very artistic though, I like the flowers and the trunks and vine and some leaves in between, I could draw and paint  a nicer one if you let me. ;)

Please feel free to pm if anyone is interested

thanks

hi guys

i m receiving more advices than real interests in the project

i mentionning it again here,we would be able to go alone on this and sell chips with better profit ,instead we have decided to share the risk and try to sell the chip at costs price to those who will invest with us.

this is not a preorder or else project but a partnership

i hope more interested people will be in it

thanks

This user has REAL interest in this project.

It could be integrate nicely with this proposal: https://bitcointalk.org/index.php?topic=294551.msg3157477#msg3157477 in a 2nd or 3rd iteration.

go on your own. others not interested. you got the money and talent why need us.

well in your project you re starting with already bought chips.

here we start from almost scratch

we still need trolls like you to make everyone happy

there is a french quote saying
il faut de tout pour faire un monde

meaning we need people like you and people like us to make it work

Why not IPO?

i didn t wanted to go this way as i find it complicate plus i was aiming to find 10/20 investors in order to get things interesting and manageable.

 but if you have some knowledge on this and preparing all the paperwork for this

why not?

The "board" we are designing is the 'from scratch' part... possibly when you are ready we can work this crowd funded chip into the design as well. We love the idea of using something that is community based and could marry into the open hardware design we are proposing.

Do you think a for profit company that would start its design from scratch today, tape out next spring and have products next summer (at the earliest) stands a good chance of earning back its NRE? I very much doubt it. If a for profit company cant, I dont see why you think a cooperative can.

Why do I doubt it? We are now at what,  10 companies working on 28nm designs? You may be shocked by the premiums they are asking for preorders, but those prices are simply a result of myopic miners and  the current low difficulty. Prices will just follow difficulty (+whatever reckless/clueless miners are willing to lose) and difficulty will follow asic pricing. Thats a feedback loop. The result is prices will plummet very very fast (as will profitability per GH). This would be true if there was just one supplier, but with a dozen that are all racing to lock in orders this will happen as fast as these companies combined can produce asics. Unless they are all even more inept than BFL, by the time you reach the market with a product, you will have to compete against 10 companies that have already recovered their NRE and are dumping chips and preorders on to a saturated market for prices barely above marginal cost, a tiny fraction of what they are today. You wont stand a chance recovering your investment.

Even if Im wrong about that, what could you possibly achieve? Assume you would be ready today, and you would be selling chips at cost (which means ~50x lower than current asic prices). What would happen? First off, you will be swamped by orders and you would never be able to fulfill them. Even if you could supply a massive amount, you would just push up difficulty tremendously, accelerating the scenario I described above. Sure, you would lower asic prices even faster than they would otherwise, but profitability would evaporate equally fast due to the D. You gain the community nothing.

There is only one "problem" right now with these asics; miners arent doing the math, or are deluded and are willing to pay far more than whats good for them. Thats the only reason prices are what they are. If miners wouldnt be willing to pay them, they would drop, simple as that. I dont see how you can fix this problem by producing yet another asic. Providing cheap asics wont solve it, it would just result in even more miners expecting windfall profits. In fact it might make the problem worse; if miners cant figure out today that $3/GH in January is a loss, how will they understand that $0.2/GH is a loss? And it will be a loss if you could supply those chips on time to everyone that wants them, since you havent solved the fundamental problem of miners overbuying, you just provide them another way to overbuy. And if you cant deliver on that demand, you will just become another BFL.

easy friend, you know they are promoting their product or group buys etc, yes not easy to make a chip, since you are almost done and have your own funding, just do it yourself.

hi puppet

i ve read briefly as it was too long,

general idea is to sell the chips to our investors at costs price but not to all customers.

our investors will have a limited supply on the first run and on the second one.

From there,they will have a discount price and would be able to act as exclusive wholesaler or whatever they wish to do with their chips.

Normal customers will pay normal price.

i do not intended to flood the market with my chips but idea is to make it affordable and viable for my investors and be able to compete on what will come available in the market.

as every plan there is as much chance that we fail than we succeed on this.

i m ready to take this risk .if there are people ready to do so it l ll be great if there is no then we ll find another way to do it.

Risk is present in every business venture so it s not for people who are waiting to make a safe investment.


i m trying to make it very transparent and i m repeating again , i m working on this since march/april but have put on hold as i ve got an agreement with bitfury and he decide not to respect his comitment  that why i ve got delayed.


thanks for reading and give me the opportunity to explain reasons ,ideas and why i m trying to do this

It is still profitable, since no one is willing to opensource the design of the chip. And I think it will be profitable for a long time...

Whats open sourcing have to do with it? Of course selling these chips now is extremely lucrative.

As for how long, it only depends on how much these companies can produce. Because they will produce as much as they can for as long as its marginally profitable, and they will sell (or use) anything they produce. Price and difficulty explosion will just be a result of those. 

How fast can this go?  For some perspective, TSMC alone does >1000 28nm wafer starts per day. One wafer equals ~70TH using hashfast's numbers. So one day of TSMC production gives 70PH worth of bitcoin asics, which when deployed,  might be enough to make it no longer profitable to produce more asics as electricity cost might exceed mining revenue at that point. There are at least a dozen other fabs.

Of course asics have to be packaged, mounted, miners assembled and shipped/deployed. Some/many companies will falter there, but it only takes one to get it right. Its entirely feasible by next summer, prices will have plummeted so far, NRE's for new designs cant realistically be recovered and by the end 2014, we might see stagnation in the network when electricity costs put a limit on the hashrate.

Let's see what will happen. Also, that means, the bitcoin becomes the new driving force for the semiconductor industry, while it is making 'great' contribution to a greenhouse environment:)   The flaw of the bitcoin system is that a lot of energy is wasted just for getting a control of inflation. Actually a better system can make better use of these energy.

Ahm, no, not really. I guess my 1000 wafer per day starts at TSMC might have been too conservative as estimate:

http://semimd.com/blog/tag/28nm/

My reckoning, bitcoin will account for not much more and possibly much less than 1000 wafers in 2014, on a capacity of almost 4 million wafers. Not gonna make a lot of headlines in silicon processing magazines.

Hopefully the mining-ASIC manufacturers will eventually realise that mining is a sucker's game, that their huge farms are obsolete, and realise how much money they missed out on by not pouring out ASICs to suckers in vast, vast quantities much much faster, driving all the conflict-of-interest manufacturers who wasted time and money setting up mining farms of their own into obsolescence...

TL;DR there is reason to sell the damn things quick instead of mining with them, because if someone does mining with them is going to be a white elephant...

(Whereas for mine-at-home folks heating their bathroom floors and heating their houses in the winter and not needing employees and not having to rent datacentre space and so on and so on, all the overheads of a commercial mining operation, mining might still actually make sense.)

Basically the normal return on investments is only maybe 5%, 10% if you are really lucky, so if you can sell the darn things at 10% profit and buy more wafers you could make 10% over and over again adding up to much more than 10% per year by the time you've churned out product for a year...

-MarkM-

While i was complaining the bitcoin mining wastes too much energy, it is true that a miner is also a heater. I guess a heater with mining capability will be welcomed in northern Europe, or people can replace heaters with mining machines...

Hi guys

the gds is ready and we are ready to tape out

we need around 1000 btc to make this happened

i ve changed the conditions

for one btc(exchange rate based ond 900 us) the crowfunder will get 1800 chips (1.25 terahash)

They will get the right to buy the same amount of chips as the same rate for their next orders.

We will work in the mean time on the pcb design but not before we get the Nre cost financed.

So let s have it done

thanks

this is the adress where to send the bitcoins

1 btc is the minimum to participate


16VKxHkM8FHnAbyMPvu24QHDRxwCKKzh2n

send your name,email adress and adress with the transactions refence to btccashcom@gmail.com

feel free to pm me

thanks

So what does this mean for a timeline? When would the chips be available. What would be the extra cost once you have the chips to get onto a pcb so that I could use?

Phil

pcb costs would be usually the same as the chip costs

so let say you ll have to add the same amount to get your pcb

i ll be able to provide more accurate number while i ll get the specs of the chip

thanks

So 1 btc now, for chips, 1 btc for PCB's in x weeks. For working boards.

And what is the number of weeks? 10, 20?

Phil

if we get the funds

it would be less then 10 weeks....

Well if you get close to your funding requirement I will put a couple of coins in, but need to see how your funding goes first.

Phil

i ll lock this thread as i ve got adjusted info from the ic house and unfortunately we ll have less chips by wafer

so i ll need to run my calculation again to get the cheapest cost possible and give every one chips at fab costs.


thanks