question regarding asic hardware

[jeprokzki]

hi guys im not new in bitcoin nor experienced im still in the middle of learning this bitcoin hardware...im planning to make a custom mining rig using custom boards not CPU/GPU can anyone recommend what to buy and where to buy? with a decent hash rate and is this already configured? like you will need to program it to use for bitcoins etc. or just use a mining software to run it instantly?

thanks in advance


sorry for my bad english...

[MaTachi]

AFAIK there aren't any proof that any one has been shipped yet. Butterfly Labs claim they are working on some.

[jeprokzki]

hi thanks for the reply but just to make it clear this. im referring for the custom made by us miners not with bfl or avalon company etc..

[MaTachi]

Making a custom rig with 10 graphics cards doesn't make it an ASIC rig. ASIC stands for "application-specific integrated circuit" and is in this case only made for hashing Bitcoins. You can by design not use such a machine for anything else. AFAIK no such machine has been shipped.

[bitchef]

Hasn't Avalon shipped a few units already?

[workin_machine]

Yeah that would really interesst me too if Avalon and BFL ever had delivered any "ASIC" products.

And i think for a normal guy the development of an ASIC-Rig is quite not realizable. These are highly specialized systems
and a lot harder to produce then a 4 graphics card mining rig.

The only persons who can build asics on their own would be some real nerd technologic students ;-)

[legendster]

To build any asic or Fpga card or boards yourself you need to be a student of Electrical Engineering. Here in India we have this topic in the 3rd Semester of our course.

From what I have gathered .. building the board is easy, its the chip that counts & isnt something that is easily made by sitting at home.

But I am interested in making my own rig as well be it a low power consuming yet high hashing FPGA board or a high hashing but useless ASIC board.




One thing is certain, that the Bitcoin community is only open these days to "old/early" adopters, as they have already made enough bitcoins when the difficulty was low. People like you & me try to get in but we cant .. basic reasons :
1. We just do not have thousands of dollars or hundreds of Bitcoins to buy us or PRE ORDER ouselves any mining rig - if we did we wudnt be here looking for easy way to make money now would we ?
2. The more number of people start mining the higher is the difficulty & lower is the reward / payout for existing miners.

Currently I have a single Gt 220 installed in my pc & it gives me a speed of 5 mh/s ~ so imagine my condition.

[BTC4Amazon]

To build any asic or Fpga card or boards yourself you need to be a student of Electrical Engineering. Here in India we have this topic in the 3rd Semester of our course.

From what I have gathered .. building the board is easy, its the chip that counts & isnt something that is easily made by sitting at home.

But I am interested in making my own rig as well be it a low power consuming yet high hashing FPGA board or a high hashing but useless ASIC board.




One thing is certain, that the Bitcoin community is only open these days to "old/early" adopters, as they have already made enough bitcoins when the difficulty was low. People like you & me try to get in but we cant .. basic reasons :
1. We just do not have thousands of dollars or hundreds of Bitcoins to buy us or PRE ORDER ouselves any mining rig - if we did we wudnt be here looking for easy way to make money now would we ?
2. The more number of people start mining the higher is the difficulty & lower is the reward / payout for existing miners.

Currently I have a single Gt 220 installed in my pc & it gives me a speed of 5 mh/s ~ so imagine my condition.

Complaining because you didn't start using bitcoins years ago isn't going to make bitcoins fly into your wallet, so if I may suggest going and buying a few so in 2+ years you can say you paid ~150 USD for something thats worth 1k


--
BTC4A

[whizz94]

Don't dabble with what you don't understand.  The goal of some ASIC manufacturers can reasonably be expected to be to get a prepayment off you, make a few ASIC's, and then run them in test for a month "just to be sure".  By the time the hardware gets posted to you, they'd be *testing* the next one, so the network hashrate will have gone up again.  You'd not lose as much as by putting money into the global financial ponzi system but to want to buy stuff solely to get rich is asking for it.

[Aahzman]

hi guys im not new in bitcoin nor experienced im still in the middle of learning this bitcoin hardware...im planning to make a custom mining rig using custom boards not CPU/GPU can anyone recommend what to buy and where to buy? with a decent hash rate and is this already configured? like you will need to program it to use for bitcoins etc. or just use a mining software to run it instantly?

thanks in advance
sorry for my bad english...

You're not talking about ASICs.  At best, you're talking about FPGA-based mining boards, and with mass-release of ASIC-based miners looming on the horizon, the FPGA ship has sailed and become (mostly) unprofitable. 

If someone were to give you a whole bunch of FPGA boards, already configured for bitcoin mining, I'd say yeah, a couple hundred bucks to assemble them into a rig might be worth it. Otherwise, by the time you buy boards, buy FPGA chips, get them configured/working, then assembled into a rig, you'll be very very deep in a financial hole that the rig will NOT mine you out of.

[stromma44]

and with mass-release of ASIC-based miners looming on the horizon

I hope your right. The situation now seems like "give us money now for a preorder, you will get your unit  sometimes (maybe)".

[Aahzman]

and with mass-release of ASIC-based miners looming on the horizon

I hope your right. The situation now seems like "give us money now for a preorder, you will get your unit  sometimes (maybe)".

Well, there is that. :-) I'm still optimistic that BFL will start shipping "soon".  Avalons are already in the hands of miners and pools, ASICMiner has working units.  So, the technology is making its way.  If BFL is a scam, it is the most elaborate long con I've ever heard of. Dozens of people, websites, chat boards, forums, videos, photos, etc etc.  If it's a scam (and I really hope it's not), they are really committed to the bit.

[bitchef]

If BFL is a scam, it is the most elaborate long con I've ever heard of. Dozens of people, websites, chat boards, forums, videos, photos, etc etc.

I think we've already seen many scams at that scale. I doubt it started as a scam, but if you are failing to meet your promises and keep accepting preorder money, that doesn't make your very honest either.

[reactor]

Bingo.  But on the plus side, even if Avalon and ASICMiner both start shipping, there is sure to be another company (likely not BFL) who jumps on the bandwagon and starts making more consumer-grade ASIC units available.  At that point, an ASIC will likely be the equivalent of a 7970 card, so you'll have to purchase and maintain several to pull a profit (after the hardware pays for itself).

It's like anything else, if you're in the first wave of ASIC's you are set to profit pretty nicely, but if the projections are right we'll all be talking about tweaking our [ASIC Vendor 4] units to get 67000GH/s instead of 63500GH/s.  Nature of all things.

[jeprokzki]

hi guys thanks for all of your replies as legendster stated.

bitcoins when the difficulty was low. People like you & me try to get in but we cant .. basic reasons :
1. We just do not have thousands of dollars or hundreds of Bitcoins to buy us or PRE ORDER ouselves any mining rig - if we did we wudnt be here looking for easy way to make money now would we ?
2. The more number of people start mining the higher is the difficulty & lower is the reward / payout for existing miners.

Currently I have a single Gt 220 installed in my pc & it gives me a speed of 5 mh/s ~ so imagine my condition.

that's the reason why im trying to learn how asic/fpga boards works and because i saw some threads posting build your own mining rig for low cost. i dont have a lot of money to pre-order bfl although i still not know if its really a legit product...not like primeasic its a scam.

@Aahzman

You're not talking about ASICs.  At best, you're talking about FPGA-based mining boards, and with mass-release of ASIC-based miners looming on the horizon, the FPGA ship has sailed and become (mostly) unprofitable. 

If someone were to give you a whole bunch of FPGA boards, already configured for bitcoin mining, I'd say yeah, a couple hundred bucks to assemble them into a rig might be worth it. Otherwise, by the time you buy boards, buy FPGA chips, get them configured/working, then assembled into a rig, you'll be very very deep in a financial hole that the rig will NOT mine you out of.

thanks for the reminder sir Aahzman maybe i will think twice before i spend money for this

[desert_beagle]

So OP, you're planning to make a custom asic miner using DIY?  Well, good luck with that because it takes alot of research and development to make things like that.  Not only you need large sums of cash but you also need a team of people who have the knowledge in fields like circuit desing, electrical/comp engineering.

OK let's pretend that you done the above.

Next comes where and how you're gonna manufacture these chips.  You need to establish contract with chip manufacturers in Asian countries (cheapest option) where they'll build those things for you because I highly doubt you have a manufacturing plant in your backyard.

Hope this helps.

[windmill]

good luck trying to out perform ASIC.

[leopard2]

by the time BFL ships ASICs, you can get consumer quantum computers... 

They will break SHA-256 anyways, so don't bother.

[jthelle]

There are a couple of ASIC designs available for sale, but no ready made chips... I think the best option at this time would be to wait for a fast SHA-256 chip to pop up on Octopart, then warm up your soldering iron

[25563673]

My company has started a project early this year in order to develop our own, high performance Bitcoin transaction processing system, because we were very dissatisfied with the situation in the market regarding actual availability of machines. At the core of it, there will be (of course) a full-custom ASIC. We are (usually) in the business of building highly optimized machines for finite elements computing, so we're not completely alien to the topic of application-specific systems. Developing an ASIC-based Bitcoin transaction processing system sure is a lot of work, but it's not rocket science, and it can be done even by a small company like us, if sufficient knowledge, skills, and funds are available.

We believe that we will see _a_lot_ of similar projects soon (even more so as the BTC exchange rate keeps breaking record heights by the day), with the effect that both the network's total hashing power and difficulty will climb to breathtaking heights. For those who already have invested heavily in such systems, the only choice will be to invest even more, and then re-invest the revenue that these systems generate, hoping to eventually get an edge over their competition by being the first to break even (thus generating more profit that can be re-invested, until nobody will be able to catch up anymore).

We will inevitably end up in an arms race which will make the Cold War look pale, with one sure winner, two possible winners, and many sure losers:

The sure winner is Mt.Gox, because they make money with every single transaction, without any need to engage in a ruinously expensive technological arms race.

A possible winner will be either those who can invest so much money that they'll be able to take the lead in terms of hashing power by using sheer mass of widely available stock hardware, or those who will get an edge by developing a technologically advanced solution, and are able to maintain that lead for a time long enough to break even and then make some serious revenue.

It also seems clear who will lose: all those "normal" miners who invested in expensive GPU-based rigs in the past and didn't break even yet, those who (re-)invested (too) late in FPGA technology (and also didn't break even yet because they were hit straight between the eyes by the rapid increase in difficulty), and of course those poor souls who made pre-order payments 6 or 8 or months ago, and didn't get _anything_ in return yet. With the expected increase in network hashing power (and difficulty) at the time when a considerable number of ASIC devices hit the market, they'll probably make less than those who (back then) invested the same amount in FPGA technology.

Bottomline: if you're about to invest (a lot) of money in Bitcoin mining hardware, and if your business model consists only of the vague hope to make some revenue from mining, better think twice. Chances are that you'll lose most of your money. People will still be able to make a lot of money with Bitcoin, however, with a different business model than "just mining". Consequently, my company doesn't spend all that money betting on becoming an ASIC technology leader, but rather on our (hopefully very clever) business model.