BitForce SC - full custom ASIC

[ice_chill]

My funds are ready to preorder.

[P4man]

@SgtSpike,
Clearly Im not getting my point across. Lets try a different approach.

Assume I have 28nm fab in my mothers basement and I just finished my asic miner. I put it up for sale, long before BFL; 100 GH, 1W, delivery in 3-6 weeks. You can order today.

Will you order one if I asked $100K ? $10K? $5K? $1K?

[Entropy-uc]

This really only seems to be a huge issue with this community.  The auto industry does this all the time, and people are more than happy to oblige.  If you're building a house, you're often required to put a large payment (percentage wise) up front and deadlines/budgets are missed all of the time.

I realize this is apples and oranges, but IMO, the mining community is pretty unique in their entitlement demands and expectations.  I really don't understand why every BFL thread needs go through 250 posts of bullshit.  If the argument is of morality or philosophy behind ASIC development (by anyone) then it deserves its own thread.  If it's to further hack and bash at BFL's business practices, there's a thread for that in Offtopic.

There is a huge difference between a deposit and 100% payment.  When I bid on my house I put 1% of the purchase price into escrow as good faith money.  

Regarding cars, this is an excellent example.  A dealer is charged the full price of a car he ordered the exact moment it rolls of the assembly line.  Not a minute before that.  If you want to purchase a custom configured car, you will need to pay a deposit with the dealer, but certainly not the full price.  I considered buying a Tesla a couple years ago and they were asking for less than 10% as a deposit at that time.

Car dealers are required to post a financial bond to help reimburse any customer deposits in case of insolvency.

[LordMord]

@SgtSpike,
Clearly Im not getting my point across. Lets try a different approach.

Assume I have 28nm fab in my mothers basement and I just finished my asic miner. I put it up for sale, long before BFL; 100 GH, 1W, delivery in 3-6 weeks. You can order today.

Will you order one if I asked $100K ? $10K? $5K? $1K?

I think we get you point. But what solution do you propose?

The only i see would be for BFL to price the ASIC at an already low price so that for newer customers the price does not have to be scaled down, while the difficulty increases. But that is BFLs decision nothing (expect not buying) we can do about that.

Also a low production capacity could make this problem less important. No idea what capacity ASICs would have.

A statment from BFL about this would be nice.

[yochdog]

I do have one question... What is exactly the point you are trying to make?

Regards,
BF Labs Inc.

Thanks for asking.  I have a few simple points.

About BFL
1.  You are pre-announcing this product before delivering your previously pre-announced product
2.  This is your second claim of delivering a product with ASIC hashing.  Your first was debunked as an FPGA as soon as product shipped.
3.  You seem to be using customer funds for purchases of product for development and general expenses.  If I am wrong about that please post a notarized statement from your escrow agent and I will apologize.

About business in general
1.  Paying in advance for development is an investment.  It is generally compensated with intellectual property, equity or loan interest.
2. Honest, viable businesses become insolvent all the time for lack of cashflow.  It is the reason for the Chapter 11 bankruptcy procedures.
3. Payments in advance for products become unsecured debts in a bankruptcy proceeding.  Unsecured debtors are among the lowest priority for recovery in these procedures and generally get nothing.  Even if the material is sitting in the shipping bay with a label on it, it isn't yours under bankruptcy.
4.  For reason #3 above it is unethical, and a violation of consumer protection laws in many places to sell product in advance of availability.  For the same reason, Visa and Mastercard require shipment before a sale can be charged.
5. Companies at risk of insolvency often make very generous guarantee offers.  A guarantee has no revenue cost in the present, and has no value in bankruptcy either.

I can't see ever being a customer of yours given your business practices.  And your question has given me the opportunity to  nicely summarize why people should use caution.  So I will leave your thread to your fans and investors.

This a very well thought-out post.  Regardless of whether you are a fan of BFL or not, these points are worth thinking about. 

[Inaba]

2.  This is your second claim of delivering a product with ASIC hashing.  Your first was debunked as an FPGA as soon as product shipped.

I don't recall them ever claiming there was ASIC hashing.  Please cite an example?  My memory certainly could be faulty though.

4.  For reason #3 above it is unethical, and a violation of consumer protection laws in many places to sell product in advance of availability.  For the same reason, Visa and Mastercard require shipment before a sale can be charged.

Please cite case law for this.  

Additionally citation needed on Mastercard/Visa issue.  Most companies I know/have dealt with (numbering in the thousands) charge first, then ship, not the other way around.  There have been a few exceptions (less than 1%).  Are you telling me literally 99% of the mail order businesses in the US are in violation of the MC/Visa agreement, and MC/Visa haven't brought down the hammer?

[SgtSpike]

Entropy - I get charged on my credit card all the time before a shipment is made.  The company I work for does the same.  I have never heard of it violating any sort of consumer protection laws.

@SgtSpike,
Clearly Im not getting my point across. Lets try a different approach.

Assume I have 28nm fab in my mothers basement and I just finished my asic miner. I put it up for sale, long before BFL; 100 GH, 1W, delivery in 3-6 weeks. You can order today.

Will you order one if I asked $100K ? $10K? $5K? $1K?

Ok, I'll bite.

At 100 GH/s with negligible electricity usage, I would stand to make nearly $10,000/month at current price/difficulty ratios.

Most reasonable people would expect a ROI of equal to or greater than the stock market (in this case, greater, since Bitcoin investments are still generally very risky).  I personally would like to see payback within 6 months, but would probably accept a payback of up to 1 year.

So, if the price is $60k, payback in 6 months, I would order.  If it is $70k=$120k, payback in 7 months, I might not.  If it is $120k+, I definitely would not.

Now, say I have the knowledge (or an educated guess) that you might drop the price from $60k to $10.  Ok, fine.  My payback period just got increased from 6 months to 3,000 years.  Whoops, should have seen that one coming.  It'll pay back, but not in my lifetime!

Ok, so the most I am willing to risk is a 6 month payback at current rates, but with the expectation that the price could drop to as low as $10, and a maximum of 3 years payback if it does.  If the price drops to $10 for 100 GH/s, I could expect 1.3 million of these boxes to be purchased, eventually, at the current price point of ~$5/BTC.  With 1.3 million, we'd have 131M GH/s, or 131.4 PH/s of network power.  The payback on these boxes would be 6 months at that price/difficulty ratio, so they'd be generating around $1.60/month worth of BTC.

Still, those 1.31 million boxes wouldn't be produced and sold in a day.  It would take months, perhaps years to get that many produced, packaged, sold, and shipped out.  I would be expecting this, and would maybe expect difficulty to double every 2 weeks at the most.  So, the first 2 weeks, I would make $5,000, the second two weeks, $2,500, third two weeks, $1,250, etc.  Basically, my revenue would virtually max out at just below $10,000 by the time my goal ROI of 6 months is up.  I'd still be making the $1.60/month extra, but that amount would be negligible compared to my initial investment.

Of course, I want to have some extra money for my time and effort after it's all said and done.  So, if I am expecting $10,000 in revenue, the maximum I would be willing to pay might be around $7,500.

Your turn.

[RoloTonyBrownTown]

Angry neighborhood bastard mod, author of DiabloMiner, founder of Diablo Mining Company, and all around Demigod of the Bitcoin community here.

Hows this: BFL sends me the smallest ASIC product they make, I receive it, and then I will everyone that not only does it exist, but it works. This should end all the trolling, and if it doesn't I will happily start modding this forum until the shit ceases.

This should solve all the problems.

Nice try. lol

I'm quite serious. I'm tired of people attacking BFL for no reason, and I will just start discussing long term depopulation plans with Theymos if this continues. Its completely up to the trolls inhabiting this thread on what happens next.

BFL sending me a unit for inspection and testing would prove it exists and works and isn't vaporware.

Better yet, send it to Inaba.  They've already worked with him.

[P4man]

I think we get you point. But what solution do you propose?

Like I said, i dont see a solution that wouldnt hurt BFL. They could make a promise to not sell more than x GH per y months, or some other form of guarantee thats linked to difficulty and that would allow miners to have some certainty of their ROI. But then what happens when a competitor appears, who is not bound to those limitations?

Another possible solution is not selling the hardware but leasing it at a variable price, ideally linked to difficulty; I guess that would need some DRM solution, not entirely unlike largecoin. But then BFL is essentially taking all the risks and just giving a part of the mining reward away, and BFL might as well mine themselves.

The only i see would be for BFL to price the ASIC at an already low price so that for newer customers the price does not have to be scaled down, while the difficulty increases. But that is BFLs decision nothing (expect not buying) we can do about that.

Which would also hurt BFLs potential profitability, probably a lot, and only delay the problem. It would be good for early adopters, but at some point even that low $/GH will begin to become too high as difficulty rises, and BFL will have to drop prices or lose sales.

Also a low production capacity could make this problem less important. No idea what capacity ASICs would have.

If they have a full wafer maskset, there is no capacity limit. None. They could literally make millions of these chips. And it could cost them less than what only a few wafers would be worth at todays price/GH.

A statment from BFL about this would be nice.

Agreed.

[DutchBrat]

A statment from BFL about this would be nice.

And a statement will come in June, so although all this speculation is fun, it's also useless

Let's just see what they say and if they can back it up with a working product

Then people can rant how BFL will be the end of Bitcoin while they are filling out the order form for a nice and shiny ASIC, as they don't want to be left out of the party 

[P4man]

Of course, I want to have some extra money for my time and effort after it's all said and done.  So, if I am expecting $10,000 in revenue, the maximum I would be willing to pay might be around $7,500.

Your turn.

Excellent, your order will be shipped in 4-6 weeks. Your order number is #101.  
Yes we already sold 100 units this month (80 to gigavps who will receive them tomorrow). We will auction off 100 more units on ebay every month, these chips cost $10 to produce, so why not.

Care to redo your math?

[nedbert9]

Entropy - I get charged on my credit card all the time before a shipment is made.  The company I work for does the same.  I have never heard of it violating any sort of consumer protection laws.

@SgtSpike,
Clearly Im not getting my point across. Lets try a different approach.

Assume I have 28nm fab in my mothers basement and I just finished my asic miner. I put it up for sale, long before BFL; 100 GH, 1W, delivery in 3-6 weeks. You can order today.

Will you order one if I asked $100K ? $10K? $5K? $1K?

Ok, I'll bite.

At 100 GH/s with negligible electricity usage, I would stand to make nearly $10,000/month at current price/difficulty ratios.

Most reasonable people would expect a ROI of equal to or greater than the stock market (in this case, greater, since Bitcoin investments are still generally very risky).  I personally would like to see payback within 6 months, but would probably accept a payback of up to 1 year.

So, if the price is $60k, payback in 6 months, I would order.  If it is $70k=$120k, payback in 7 months, I might not.  If it is $120k+, I definitely would not.

Now, say I have the knowledge (or an educated guess) that you might drop the price from $60k to $10.  Ok, fine.  My payback period just got increased from 6 months to 3,000 years.  Whoops, should have seen that one coming.  It'll pay back, but not in my lifetime!

Ok, so the most I am willing to risk is a 6 month payback at current rates, but with the expectation that the price could drop to as low as $10, and a maximum of 3 years payback if it does.  If the price drops to $10 for 100 GH/s, I could expect 1.3 million of these boxes to be purchased, eventually, at the current price point of ~$5/BTC.  With 1.3 million, we'd have 131M GH/s, or 131.4 PH/s of network power.  The payback on these boxes would be 6 months at that price/difficulty ratio, so they'd be generating around $1.60/month worth of BTC.

Still, those 1.31 million boxes wouldn't be produced and sold in a day.  It would take months, perhaps years to get that many produced, packaged, sold, and shipped out.  I would be expecting this, and would maybe expect difficulty to double every 2 weeks at the most.  So, the first 2 weeks, I would make $5,000, the second two weeks, $2,500, third two weeks, $1,250, etc.  Basically, my revenue would virtually max out at just below $10,000 by the time my goal ROI of 6 months is up.  I'd still be making the $1.60/month extra, but that amount would be negligible compared to my initial investment.

Of course, I want to have some extra money for my time and effort after it's all said and done.  So, if I am expecting $10,000 in revenue, the maximum I would be willing to pay might be around $7,500.

Your turn.

Getting some good comments.  This one especially.  I am seriously concerned about the impact of low cost ASIC solutions on mining.  The easier the entry point the more distributed mining income will become.  And at that point will mining be a good business or will we return to the dynamics present with CPU based mining minus the low difficulty ( ease of entry, ease of operation, low income ).

Although, if mining operations become so trivial in sense of TCO, but with trivial income who will continue mining and would we see sharp decline in difficulty?

edit:  Seems to me BFL's best move would be to keep ASIC's to themselves.  Trivial mining for them, status quo for all else.

[Entropy-uc]

2.  This is your second claim of delivering a product with ASIC hashing.  Your first was debunked as an FPGA as soon as product shipped.

I don't recall them ever claiming there was ASIC hashing.  Please cite an example?  My memory certainly could be faulty though.

4.  For reason #3 above it is unethical, and a violation of consumer protection laws in many places to sell product in advance of availability.  For the same reason, Visa and Mastercard require shipment before a sale can be charged.

Please cite case law for this.  

Additionally citation needed on Mastercard/Visa issue.  Most companies I know/have dealt with (numbering in the thousands) charge first, then ship, not the other way around.  There have been a few exceptions (less than 1%).  Are you telling me literally 99% of the mail order businesses in the US are in violation of the MC/Visa agreement, and MC/Visa haven't brought down the hammer?

ngzhang did the reverse engineering that showed the single has an FPGA.  The thread is the source of my understanding that BFL claimed to be using an ASIC.

Visa operating rules are linked below.  Your experience is common.  I have on more than one occasion successfully collected a refund from a reluctant merchant by citing their failure to ship as a violation of Visa rules.  Charging the same day you ship is a far different matter, and overlooked by the card companies.
http://usa.visa.com/merchants/operations/op_regulations.html

I'm not interested in researching case law for you.  It varies by state and I have lived in some jurisdictions where the law was as I stated.  If I had a dog in this fight I would hire a consumer rights lawyer in the capital city of BFL's state to investigate and possibly file a complaint.  If you really want to know, google or make some phone calls.

[nbtcminer]

As someone who only jumped on the BFL bandwagon earlier this year (and not late last year), BFL is likely to deliver on their promise of an ASIC device that is under $5000. When they deliver and the actual specification is any best guess at the moment. Based on other hearsay / non-proven numbers in other threads we're looking at 8 - 10ghs per device (if they sell them like singles) at anywhere from $800 - 1200 (*These number are based on what current products are sell for and where their competitors products are currently priced / spec at).

As for the effects of an ASIC on the bitcoin economic and mining environment, I see the ASIC phase as being an important part next step in the development of bitcoins. The speed increase in the network may be necessary to keep up with global demand / growth with regards to the halving of rewards for mining later this year. People who are worried about their current investment in certain FGPAs or GPUs should note that this was always an inevitable step. Just as GPU mining killed off CPU mining, so will sASIC/multiple FGPAs and ASICs dominate the next mining phase of bitcoin mining. It is inevitable.

As for Inaba, he deserve more respect than some of the people here give him. I'm not being a fan boy for no particular reason; he's actually helped and contributed to this community both to his own success / detriment. People were lobbing shit at him for being a BFL "puppet" when he was just trying to get his hands on one of their products and then sharing his experiences here on the forums. He not only answered my private messages when I was lower count poster but even offered to buy my reservation spot for a BFL single when I wasn't confident enough about them delivering an actual hashing product. In closing, let’s just all be a little more civil and get back to the topic at hand: how should we as a community deal with the emergence of ASIC technology in terms of mining and can we make it just as successful as the last big jump forward with GPUs.


Cheers,

Nbtcminer

[Unacceptable]

Well,CPU's were used to mine first,then came GPU's.I'm sure there were some people who said this was a bad thing.

Then FPGA's started doing the same thing,making GPU's obsolete.Of course it would take quite awhile before they are not worth using,i.e. diff going up becuase of FPGA's.

Now ASIC's are really going put a hurt to those using GPU's,which are what the majority of miners are using.

I think this is where the arguments are coming from,with all the investments in GPU farms going to be worthless before years end.Thats alot of money & time invested going to waste.

It's a shame,but it's called progress................I believe the diff will go up,but as GPU farms start shutting down the diff will level off,where? I have no idea.......

[LordMord]

Well,CPU's were used to mine first,then came GPU's.I'm sure there were some people who said this was a bad thing.

Then FPGA's started doing the same thing,making GPU's obsolete.Of course it would take quite awhile before they are not worth using,i.e. diff going up becuase of FPGA's.

Now ASIC's are really going put a hurt to those using GPU's,which are what the majority of miners are using.

I think this is where the arguments are coming from,with all the investments in GPU farms going to be worthless before years end.Thats alot of money & time invested going to waste.

It's a shame,but it's called progress................I believe the diff will go up,but as GPU farms start shutting down the diff will level off,where? I have no idea.......

The problem with ASIC it that you can build them really cheap. But sell them at a high price until the difficulty goes up. You keep lowering the price as long as difficutly rises until you get close to breakeven.

The bitcoin economy won't suffer but individual miners will most likely. And the biggest winner is propably the ASIC manufacutere.

And the shutting of GPU famrs will provide some sort of buffer but not for to long.

[P4man]

The problem with ASIC it that you can build them really cheap. But sell them at a high price until the difficulty goes up. You keep lowering the price as long as difficutly rises until you get close to breakeven.

The bitcoin economy won't suffer but individual miners will most likely. And the biggest winner is propably the ASIC manufacutere.

And the shutting of GPU famrs will provide some sort of buffer but not for to long.

Thank god, someone who understands!
Well, the ones that dont and go for broke buying those asics at whatever price BFL decides upon, I guess will understand too - eventually.

[SgtSpike]

Of course, I want to have some extra money for my time and effort after it's all said and done.  So, if I am expecting $10,000 in revenue, the maximum I would be willing to pay might be around $7,500.

Your turn.

Excellent, your order will be shipped in 4-6 weeks. Your order number is #101.  
Yes we already sold 100 units this month (80 to gigavps who will receive them tomorrow). We will auction off 100 more units on ebay every month, these chips cost $10 to produce, so why not.

Care to redo your math?

Then I just got burned because I didn't do due diligence research on finding out how many others of these ASICs were produced.

However, I understand your point.  The concern is that ASIC mining will increase difficulty so vastly and so quickly, that miners will not be able to accurately formulate a proper ROI estimate, even if being conservative, and overestimate what they will actually be able to produce.

Certainly, you've exaggerated the scale with 100 GH/s @ 1w, but the electric usage will still be nearly negligible compared to the initial outlay for the units.

I think more accurate calculations and assumptions can be made when BFL makes their announcement.  Some people will be willing to take the risk, and others will not.  If they release 1 GH/s units, it would be much more difficult to increase the difficulty quickly compared to 100 GH/s units.  They'd have to ship 100x as many units that way.  So the GH/s per unit is an important factor, as well as the (guesstimated) price per unit, guesstimated production capacity, etc.

[nedbert9]

Of course, I want to have some extra money for my time and effort after it's all said and done.  So, if I am expecting $10,000 in revenue, the maximum I would be willing to pay might be around $7,500.

Your turn.

Excellent, your order will be shipped in 4-6 weeks. Your order number is #101.  
Yes we already sold 100 units this month (80 to gigavps who will receive them tomorrow). We will auction off 100 more units on ebay every month, these chips cost $10 to produce, so why not.

Care to redo your math?

Then I just got burned because I didn't do due diligence research on finding out how many others of these ASICs were produced.

However, I understand your point.  The concern is that ASIC mining will increase difficulty so vastly and so quickly, that miners will not be able to accurately formulate a proper ROI estimate, even if being conservative, and overestimate what they will actually be able to produce.

Certainly, you've exaggerated the scale with 100 GH/s @ 1w, but the electric usage will still be nearly negligible compared to the initial outlay for the units.

I think more accurate calculations and assumptions can be made when BFL makes their announcement.  Some people will be willing to take the risk, and others will not.  If they release 1 GH/s units, it would be much more difficult to increase the difficulty quickly compared to 100 GH/s units.  They'd have to ship 100x as many units that way.  So the GH/s per unit is an important factor, as well as the (guesstimated) price per unit, guesstimated production capacity, etc.

All of this is true and it's totally pissing me off due to making Bitcoin mining that much more risky.

[JWU42]

So finally got a note back from Jody that there isn't an NDA for the customers - it related to manufacturing partners and their VCs...    

Communication breakdown...

EDIT - Sorry for the dupe thread - just making my way through the 6 pages from today to see that there was an earlier post...