BitForce SC - full custom ASIC

[Turbor]

I love the panic here

[1714490283]

1714490283

[1714490283]

1714490283

[1714490283]

1714490283

[1714490283]

1714490283

[1714490283]

1714490283

[ngzhang]

Can we order now?

Didn't take you long, did it?

No, seriously, this pre-announcement will pull the rug out from underneath all other vendors - the Icarus/Lancelot guy in mainland China, Stefan in Germany, the British guys, the American guys with their 6500 board, even the genius-level bitstream guys like EldenTyrell and Bitfury.

It's going to be a bloodbath.

personally, i think the " bloodbath"  "should" started around last year. a 1.05G/20W/499$ device "should" kill all of us, as you listed.

but we stand here before you now, truthfully unafraid.

[bitfury]

Can we order now?

Didn't take you long, did it?

No, seriously, this pre-announcement will pull the rug out from underneath all other vendors - the Icarus/Lancelot guy in mainland China, Stefan in Germany, the British guys, the American guys with their 6500 board, even the genius-level bitstream guys like EldenTyrell and Bitfury.

It's going to be a bloodbath.

personally, i think the " bloodbath"  "should" started around last year. a 1.05G/20W/499$ device "should" kill all of us, as you listed.

but we stand here before you now, truthfully unafraid.

No no, the real bloodbath will start when there will appear ASIC not supported by community... Then investors will make bloodbath to such ASIC producer - because hashing ASIC would be worthless... So there will be collective motion towards hashing algorithm change... Remember, users first, miners next... BitCoin was not for miners... There were multiple clones "for miners", and where are they now ? ;-))))

[Entropy-uc]

No, seriously, this pre-announcement will pull the rug out from underneath all other vendors - the Icarus/Lancelot guy in mainland China, Stefan in Germany, the British guys, the American guys with their 6500 board, even the genius-level bitstream guys like EldenTyrell and Bitfury.

It's going to be a bloodbath.

I am certain that is the general effect they are looking for with this particular announcement.  Let's consider some facts, and how they might pertain to BFL's business model.

1)  Enterpoint, a professional FPGA design and build house enters the market with introductory pricing competitive with the single.  They communicate openly and credibly on the forums, report progress and problems, and are keeping to their promised schedule so far.

Q1)  How does this impact the rate of new orders for BFL singles?

2) Enterpoint does not require any payment until product is ready to ship.

Q2) How does this affect orders for any BFL product when they are expecting payment up front with indefinite shipping times? Will customers continue to be willing to supply BFL with their working capital?  (that means being an uncompensated investor since without escrow, their funds are being used for general operations)

3) Eldentyrell promises an announcement on his custom code for Spartans.  Initial claims are 25%-50% faster than current bitstreams.  The single will not benefit as they use a different FPGA.  At the most optimistic levels, cost per hash on shipping quad boards could be lower than mini-rigs that still are not shipping.

Q3) How does this affect new orders for singles and mini-rigs?

Now as an exercise for the reader; imagine that you are running a company that relies on the cash from your order book to continue operating.  Now imagine that the rate that new orders arrive suddenly declines.  As you ship product off your order books you are losing working capital.  Eventually this will result in a negative cash balance at your bank, even though you have a functioning business model.

What would you do?

Would you perhaps announce an imaginary future product even though you haven't shipped your previously announced product?  Would you offer a trade in guarantee on existing products to try to refresh your order flow?

Personally, I would find investors to properly capitalize my business.  Others might choose a path that exposes their customers to the risk of complete losses on their orders.

[P4man]

This is not how we have envisioned our production. The strategy you have mentioned is effectively against
the favor of early customers and in favor of late customers. Our strategy has always been in favor of our
early customers.

No matter how you envision it, its the reality of ASIC development: (very) high NRE with nearly negligible per chip, and therefore per GH, costs. Since the bitcoin mining market is mostly limited by how much bitcoins are worth, you get a very weird situation.

lets put it this way; a processed wafer costs on the order of a 1-2000 dollar and should provide on the order of a terrahash or more worth of asic hashing power; actual numbers dont really matter here. Ignoring the other costs (PCBs etc ) which are relatively minor,  a single batch of 12 $1000 wafers could potentially provide as much hashing power as the entire network, and therefore be worth millions today.

But if you sell the equivalent of hundreds of wafers, its not going to net you hundreds of millions, the market just isnt there for that, it will still net you on the order a few million dollar.  So if you can generate a comparable revenue by selling several dozens of devices or several thousands of them, which makes most sense?  You go with the former obviously, anyone would. Not that it would prevent you from selling thousands later on.

So anyone buying that early terrahash of asic power would get what will probably seem like a great deal; yet he could lose most of his investment 6 months later just because you or a competitor decided to crank up the production volume almost arbitrarily. Suddenly his 1TH is no longer 5 or 10% of the network but 0.05% and it will only take a relatively minor investment by you or a competitor to produce those extra chips.

So this is not criticism of BFL, its just the logical consequence of a bitcoin ASIC that can generate virtually unlimited hashrates while the total profit miners can generate is fixed. As a result, a miner his ROI is completely dependent on your, or a future competitors production volume. Knowing the variable cost of production is virtually nothing, thats pretty damn risky. So I seriously think some sort of assurance that you will not produce more than x TH over y time would be a very good idea to reassure early customers.

[ice_chill]

Can we order now?

Didn't take you long, did it?

No, seriously, this pre-announcement will pull the rug out from underneath all other vendors - the Icarus/Lancelot guy in mainland China, Stefan in Germany, the British guys, the American guys with their 6500 board, even the genius-level bitstream guys like EldenTyrell and Bitfury.

It's going to be a bloodbath.

personally, i think the " bloodbath"  "should" started around last year. a 1.05G/20W/499$ device "should" kill all of us, as you listed.

but we stand here before you now, truthfully unafraid.

No no, the real bloodbath will start when there will appear ASIC not supported by community... Then investors will make bloodbath to such ASIC producer - because hashing ASIC would be worthless... So there will be collective motion towards hashing algorithm change... Remember, users first, miners next... BitCoin was not for miners... There were multiple clones "for miners", and where are they now ? ;-))))

You are saying USERS and MINERS as being different yet are putting them in the same category, your statement contradicts itself.

If users are put first, then network security is a priority, and user meaning a person who uses the currency to buy and sell. Users don't care whether the transaction is verified by CPU, GPU, FPGA or ASIC.

If miners are put first, then it would mean keeping the code unchanged so the ASIC does not become obsolete.

Changing the code means users get reduced network security by taking ASICS out of action. So no-one benefits.

[Epoch]

personally, i think the " bloodbath"  "should" started around last year. a 1.05G/20W/499$ device "should" kill all of us, as you listed.

but we stand here before you now, truthfully unafraid.

I think what ngzhang is saying is "seeing is believing". One cannot dispute the fact that a 1.05G/20W/$499 device was (pre)announced, and in the end turned out to be a not-quite-as-svelte 0.832G/80W/$599 device.

I have been disappointed by the Singles' real specs relative to the inflated hype and anticipation, but have purchased them in spite of that and in spite of the delays and have been more than satisfied.

So based on that history, for the upcoming (pre)announcement of the SC, I'll remain cautiously optimistic but take any announced specs and shipping dates with the prerequisite grain of salt. Looking forward to more detail.

[Turbor]

In a fast moving world, today's kings are tomorrow's fools

[bitfury]

Can we order now?

Didn't take you long, did it?

No, seriously, this pre-announcement will pull the rug out from underneath all other vendors - the Icarus/Lancelot guy in mainland China, Stefan in Germany, the British guys, the American guys with their 6500 board, even the genius-level bitstream guys like EldenTyrell and Bitfury.

It's going to be a bloodbath.

personally, i think the " bloodbath"  "should" started around last year. a 1.05G/20W/499$ device "should" kill all of us, as you listed.

but we stand here before you now, truthfully unafraid.

No no, the real bloodbath will start when there will appear ASIC not supported by community... Then investors will make bloodbath to such ASIC producer - because hashing ASIC would be worthless... So there will be collective motion towards hashing algorithm change... Remember, users first, miners next... BitCoin was not for miners... There were multiple clones "for miners", and where are they now ? ;-))))

You are saying USERS and MINERS as being different yet are putting them in the same category, your statement contradicts itself.

If users are put first, then network security is a priority, and user meaning a person who uses the currency to buy and sell. Users don't care whether the transaction is verified by CPU, GPU, FPGA or ASIC.

If miners are put first, then it would mean keeping the code unchanged so the ASIC does not become obsolete.

Changing the code means users get reduced network security by taking ASICS out of action. So no-one benefits.

There's no contradiction. It is are users, who are willing to use BitCoin who by-fact pay compensation to miners for the network security. Not vice-versa.

So open-asic would work, then everybody will know exact takeover cost, while NDA-ASIC won't work. So BFL would likely end up in condition, where they would have to sell these chips very close to silicon cost, and opening specs OR OTHERWISE - they would become private owners of overall BitCoin system security. As it was correctly said, that if you count cost of silicon wafers - that are trivial costs to outperform current network hashing power.

Then - when someone privatizes BitCoin - it may be good or not... For example - if I would privatize it, and control majority, what I would do ? Ask yourself that question. Because actually I will become LIABLE for all damages that users could get by my actions, but of course there's no legal framework about it. And if I start with mining only purpose - then - I will be screwed - because nobody wants "just mining", people need goods and services paid in secure way. So model would change from trust-nobody decentralized to trust-one who has majority, not much difference compared to webmoney then. And whether it will work or not - it depends on actions of that de-facto owner.

And then it comes to BFL.......... and trust..... would I trust my assets to them with their approach ? Ask this question to yourself... And ask another question - who you know here on forum to whom you would trust say $10k holding in the system, and knowing that he say can dump $100M cap to $50M for own profit on exchange rates and won't do that ?

What I definitely know that greedy people would screw this thing likely, as they will not pass any opportunity to make money, even at delicate moments, where making money puts risks on all users... Even worse if they will silently start ripping off users... Then bankers would just laugh at it and say - damn internet crooks... So that 20W / 1Gh/s and other "pre-announce" buzz definitely shows how you can trust your "rulers of BitCoins" :-) Banking / payment systems is all about trust and then people behavior...

[spiccioli]

personally, i think the " bloodbath"  "should" started around last year. a 1.05G/20W/499$ device "should" kill all of us, as you listed.

but we stand here before you now, truthfully unafraid.

I think what ngzhang is saying is "seeing is believing". One cannot dispute the fact that a 1.05G/20W/$499 device was (pre)announced, and in the end turned out to be a 0.832G/80W/$599 device. Quite a different beast. In some markets, specific Spartan offerings remain competitive with that.

I think ASICs will destroy bitcoin as we know it, let's try to explain...

if 100/200 units equal 10 TH/s which is current network hashing power and if they are delivered before year end at year end every GPU and a good number of FPGAs will be out of business or near to be out of business.

Now at year end block reward halves, so ROI edit: time of every one, ASICs included, doubles (I don't see BTC value doubling because of block reward halving).

If ASIC producers keep selling their units every unit more makes ROI of every other unit even longer to the point they have to lower unit price to be able to sell them but doing so they make first ASIC buyers ROI even longer...

I think ASICs arrived too soon, bitcoin market is not big enough for them.

ASICs would be ok on a growing market which lives on transaction fees, where every year transactions double and so difficulty can double as well to the point where you _need_ an ASIC to be able to hash and get some reward.

Right now ASICs are the tragedy of the commons unfolding...

spiccioli

[DiabloD3]

Right now ASICs are the tragedy of the commons unfolding...

DUN DUN DUNNNNNNN

[DeLorean]

Can you give us any indication of price? >$5,000? >$25,000?

[SgtSpike]

I am a happy BFL customer.

2 questions:
1)  Where can I sign an NDA to get more information?
2)  Can I order yet?

[BFL-Engineer]

Can you give us any indication of price? >$5,000? >$25,000?

Can't give you any exact numbers yet. But for a certain model, both numbers are too big,
and for another, probably both are still big... we have to decide, but it won't be unaffordable...


Regards,
BF Labs Inc

[DeLorean]

Can you give us any indication of price? >$5,000? >$25,000?

Can't give you any exact numbers yet. But for a certain model, both numbers are too big,
and for another, probably both are still big... we have to decide, but it won't be unaffordable...


Regards,
BF Labs Inc

Thanks, just looking for a ballpark figure.

[BFL-Engineer]

I am a happy BFL customer.

2 questions:
1)  Where can I sign an NDA to get more information?
2)  Can I order yet?

For NDA you need to contact office@butterflylabs.com. For the order part, well if you already
own any singles or have any mini-rigs in order, you already have the credit for trade-in.
Please stand by for official announcement for ordering...


Regards,
BF Labs Inc.

[ice_chill]

I am a happy BFL customer.

2 questions:
1)  Where can I sign an NDA to get more information?
2)  Can I order yet?

For NDA you need to contact office@butterflylabs.com. For the order part, well if you already
own any singles or have any mini-rigs in order, you already have the credit for trade-in.
Please stand by for official announcement for ordering...


Regards,
BF Labs Inc.

Music to my ears

[norulezapply]

I placed an order for a MiniRig just 2 weeks back. Talk about timing 

2 weeks? try 2 days (not BFL products though)

watching...

[RoloTonyBrownTown]

Well done BFL, glad to hear you're still innovating.   Sorry to see the same sad old trolls up to their own tricks (you'd think, after they were humiliated when you came through the first time around, they'd have learnt their lesson, but there's no accounting for stupidity I suppose).

As for the competitors getting scared and starting attacking BFL, I guess you're scared for a reason hey.   Here's an idea, let your product stand on its own merits without the underhanded attacks. 

[MrTeal]

Just to clarify your earlier statements, the trade in for Singles/Minirig buyers is only for the original buyers and not people who have purchased BFL Singles second hand, correct?