1GH/s, 20w, $700 (was $500) — Butterflylabs, is it for real? (Part 2)

[CubedRoot]

Let us pray they don't literally ship a brick ... or at least something that will turn into a brick in 4 months etc.

Some scammers get away with shipping a brick and claiming to have proof of delivery too and the buyer loses out.

Well the earliest pre-orders are now almost 10 weeks old so BFL could ship nothing just send a "you got fucked" email and the buyer would have no recourse.  Paypal chargeback window is long since expired.

Thats also a very biased statement as BFL said they would gladly issue refunds to those that wanted, plus if anyone was skeptical they should have withdrawn long before the window closed.

[DeathAndTaxes]

Thats also a very biased statement as BFL said they would gladly issue refunds to those that wanted

Well they also said they would ship product a month ago, that it would consume 19.8W and that it would operate at 1.05 GH/s.  A company can say a lot of things.  The point isn't that BFL will take the money it is just that they don't need to ship a brick to walk away with the money, they money is already theirs. Those funds are theirs and it just comes down to trust.

plus if anyone was skeptical they should have withdrawn long before the window closed.

I love this sentiment.  Someone who placed an order in good faith should have known the company was making claims it couldn't meet.  If they didn't, were to slow cancelling, or were just too trusting it is their fault not the company making false promises.

[BFL]

The point isn't that BFL will take the money it is just that they don't need to ship a brick to walk away with the money, they money is already theirs. Those funds are theirs and it just comes down to trust.

Hi,

I wonder what it is that you think we are here at Butterfly.  Why would you assume there is a reason to distrust?  

I don't expect you to root for us, but it seems you've taken a sour angle with your repeated suggestions that we're not acting in good faith.  Yes, we made an error in power consumption estimates based on our simulation analysis but I think you know enough to understand these figures are usually a pretty good guide, so using them as our performance expectation isn't an act of malice as you suggest...  The reality is mining is a unique application to cryptography hardware and we've had to learn that wrinkle.  Nothing nefarious here.  

Furthermore, in consideration for the unknown we set an open guarantee to our customers that if we failed to meet performance or delivery estimates, we'd accept purchase cancellations with a full refund.  This was all done in good faith and we've followed through without fail while keeping our customers informed as development and manufacture worked it's way through the process.  We've demonstrated our product in the flesh to a forum member of requisite knowledge and esteem.  All things considered, he seems pretty impressed and he's reported as much.  We're now adjusting for increased power consumption which has added a few weeks to the production cycle and we're on track for product delivery in January.

Looking at this from my perspective, I just don't quite understand why you (and three others) choose to see all this in an ugly way.  I'm left to wonder why.  Are you just a forum troll who's found an opportunity to flash his great insight?  Are you secretly a disgruntled competitor?  Are you offended when a new product isn't shrink wrapped with a consumer reports stat sheet before it crosses your eye?  Are we a threat to you in some way?  Honestly, I just don't get it.

Maybe this should have been a PM, but I would appreciate an answer to the puzzle and I expect others might want to know as well.

Regards,
BFL

PS.  Merry Christmas

[Syke]

Why would you assume there is a reason to distrust?

The animosity you're seeing on the forum is due to your lack of honesty. Bitcoiners don't like seeing fellow members scammed. You set off a lot of red flags, and rightly so. Your actual product is miles away from what you claimed it would be. You're still not answering repeated requests to disclose your chips so that an independant evaluation of your claims may be made. Your actions are causing the distrust.

Let me ask you a very direct question. When was the first time you measured the actual power draw from an actual unit?

[Inaba]

Sorry, but to say the product is miles away from what it was claimed is patently ridiculous.  Even at the power draw and hashrate I saw, which is IMHO a worst case scenario, the product is somewhat off.  It's not "miles away" by any stretch of the imagination.  20w vs 80w.  Yes, a factor of 4.  So is 1w vs 4w, but that doesn't mean it's a huge amount in absolute terms.  Whether it's 20w or 80w, it's still better by a factor of 6 (being generous here, it's more like a factor of 7 or 8) to any GPU solution and at LEAST 3x possibly 4x better than any other FPGA solution... So trying to couch it in terms of "factors" on how off the power estimates were is disingenous.  Put it in terms of absolute values.  It's 80w (and likely the shipping version will be more efficient), compared to 20w.  I'll take the 60w increase in place of my power hungry GPUs screaming in my basement, thanks.  And personally, I'm not going to quibble over 170 - 200 MH/s as a worst case scenario, and likely less than that as well.  Would I like a 1.05 GH/s at 20w?  Hell yes, but the reality and the initial target are anything but "miles away" from each other.  

Fuck seriously.  I am trying to stay out of this, but the rampant hate and complete bullshit some of you are spewing are borderline pathological and far, far over the border of logical or reasonable.  I see mistakes and misunderstandings, but I see little if any outright dishonestly.  But I'll tell you what I do see, is a whole bucketload of a complete lack of understanding or experience in design, business operations, marketing and development.  Anyone who's actually done those things understands exactly what's happened so far and why... anyone who hasn't really shouldn't be speaking.

[BFL]

The animosity you're seeing on the forum is due to your lack of honesty.

You see, Sky...  that's exactly what I mean.  There has been no dishonesty.  Not satisfying your curiosity by keeping the nature of our technology private isn't being dishonest.  What's relevant is performance and as far as that goes, we've been perfectly open and have shared the details along the way with both the public and our customers.       (Answering your timing question...  about two weeks prior to public demo.)

Regards,
BFL

[BFL]

The animosity you're seeing on the forum is due to your lack of honesty.

You see, Sky...  that's exactly what I mean.  There has been no dishonesty.  Not satisfying your curiosity by keeping the nature of our technology private isn't being dishonest.  What's relevant is performance and as far as that goes, we've been perfectly open and have shared the details along the way with both the public and our customers.       (Answering your timing question...  about two weeks prior to public demo.)

Regards,
BFL

Yet you neglected to answer the question he asked in that same post. I'm not saying you're illegitimate, you just seem to be good at damning yourself.

Read it again and you'll see the answer clearly stated.   " (Answering your timing question...  about two weeks prior to public demo.)"

[rjk]

Yet you neglected to answer the question he asked in that same post. I'm not saying you're illegitimate, you just seem to be good at damning yourself.

(Answering your timing question...  about two weeks prior to public demo.)

Reading comprehension much? lol...

[Syke]

The animosity you're seeing on the forum is due to your lack of honesty.

(Answering your timing question...  about two weeks prior to public demo.)

Then why didn't you come clean at the time that your measurements were off by a factor of 4+?

And I hardly call letting one person see the machine a "public demo".

[legolouman]

Oh my, I have failed hard. I'm going to go cry in a corner. While crying in a corner, I'd like to ask, why you didn't you tell us?

[BFL]

The animosity you're seeing on the forum is due to your lack of honesty.

(Answering your timing question...  about two weeks prior to public demo.)

Then why didn't you come clean at the time that your measurements were off by a factor of 4+?

And I hardly call letting one person see the machine a "public demo".

Obviously, we did.  We demonstrated shortly thereafter.  Where's the conspiracy?  

Honestly...  at this point I feel embarrassed for taking the time to respond at all.  I think we've got a pretty good product, but if you don't want to buy it then you don't have to.  Really.  It's up to you...  you're not being forced into anything.  Just relax.

Good night.  I'm off to bed.  Merry Christmas to all.

BFL

[DeathAndTaxes]

The point isn't that BFL will take the money it is just that they don't need to ship a brick to walk away with the money, they money is already theirs. Those funds are theirs and it just comes down to trust.

Hi,

I wonder what it is that you think we are here at Butterfly.  Why would you assume there is a reason to distrust? 

Because you have done nothing to earn any trust.

Did you ever have a prototype running @ 1.05GH and 19.8W?
If not when did you realize your power estimates were off by a factor 500%?
When was your first prototype built?  A month ago? two? three?  How long have you known there is no possible way you could get those power claims?

Do you think it is trustworthy to offer "GUARANTEED PERFORMANCE" and never mention the specs were based on simulations (and everyone knows ISE analysis is a shot in the dark) and you had no physical product anywhere near that spec?

Say the day prior to demo w/ Inaba did you even try running the mining code?  What performance did you get?  What power consumption?
Why didn't you update your website?  Why did you update your blog or this forum?  It took Inaba posting the REAL metrics before you made any changes to your claims?  Why?  Do you think that inspires trust?

Care to comment on this:

The BitForce processor card is a proprietary implementation of both FPGA and ASIC technology.

The performance of your hardware doesn't come from any ASIC technology does it?  This claim is simply double speak.  The USB controller is an ASIC so technically any FPGA board is a proprietary implementation of both FPGA and ASIC technology.  It is this kind of

On same page you might want to update this:

4. What is the power consumption of your BitForce Single?
The BitForce SHA256 processor has been built for extremely low power consumption. The exact power draw will depend on the end user application.  However, at full use, power draw may achieve a maximum draw of 19.8w

Do you believe your rig box will get 20 MH/W?  Is this based on any actual product (even a single physical board) or more untested simulations? If not why are you still advertising it as such?  Will that be another mistake too.  It will only be updated on demo day?

Rigbox: 50.4 Giga Hash / S @ 2,500w

Everything from your company has been half truths and misinformation.  You haven't been open or transparent about anything.  Trust is earned.  Here is the crazy part ... I may even by FPGA from you in the future when I trust I wouldn't be sold a bunch of hot air.  Your personal attacks against me aside (BTW to answer your inquiries I own no FPGAs, I have no FPGA products, and I have been on this forum longer than your company has existed ).

I think the response would have been complete different if you had launched a site WITHOUT PRE-ORDERS came to the community and said "hey we
are working on a project to get FPGA below $1 per MH is there any interest".  Some involvement other than a bunch of overpromising and under-delivering.  Promise less, delivery more, be transparent and open.  That builds trust.

Hey it isn't personal man, I may even buy an FPGA from you someday.

[Syke]

at this point I feel embarrassed for taking the time to respond at all.

Yes, it is quite embarrasing for you to be caught in your lies.

[btc_artist]

Good night.  I'm off to bed.  Merry Christmas to all.

Thanks for sticking around and answering questions. Merry Christmas to you too!

[DeathAndTaxes]

20w vs 80w.  Yes, a factor of 4.  So is 1w vs 4w, but that doesn't mean it's a huge amount in absolute terms.

Of course it does.  What matters is performance per watt.  It is like saying CFL don't save much power because it is 12W vs 60W see less than a 50W difference.  The reality is the CFL uses 80% less power the change in performance per watt is significant.  To replace your 15GH would require ~20 units.  That is 1200W more load.  That's $1000 to $2000 a year higher energy costs.  

Whether it's 20w or 80w, it's still better by a factor of 6 (being generous here, it's more like a factor of 7 or to any GPU solution and at LEAST 3x possibly 4x better than any other FPGA solution...

Not sure how you reached that conclusion.  It is 10MH/W.  

To put it into perspective:
GPU can get 2.5MH/W is done right.  
You "could" underclock and undervolt to push that to >3MH/W.  
If 7800 series use VLIW5 they will be getting almost 4MH/W stock.  
Current get FPGA are getting up to 22MH/W.  
28nm FPGA will likely push that up to 40MH/W range.  

[fred0]

20w vs 80w.  Yes, a factor of 4.  So is 1w vs 4w, but that doesn't mean it's a huge amount in absolute terms.

Of course it does.  What matters is performance per watt.  

Whether it's 20w or 80w, it's still better by a factor of 6 (being generous here, it's more like a factor of 7 or to any GPU solution and at LEAST 3x possibly 4x better than any other FPGA solution...

Not sure how you reached that conclusion.  It is 10MH/W.  

I think that Inaba reached that conclusion by estimating profitability, not power efficiency.  If power efficiency is your main criteria then the ztex is the proper choice for you.

Currently the BFL widget has the shortest break even period, I think that may be the most important criteria for most other miners.

In any case, I included the links so you can customize the estimates for your particular operation.
ztex x10 units for volume discount
Icarus x5 units
X6500 x10 units for volume discount
BFL widget x2 units

[Edit:
Please note that for all links, the "break even after" assumes the block reward does not change for the entire break even period.  However, for all but the BFL widget, the breakeven period is greater than 500 days (at the current bitcoin price).  A more accurate projection would require recalculating at the time the block reward halves, while adjusting hardware cost based on what has been paid.

I will leave this task to those miners who are so inclined.

Also, market conditions like difficulty and price are not addressed.  I am not worried about difficulty(since greater difficulty follows greater price), but price can have a profound impact on the payback period.
]

[Inaba]

Yes, my conclusions are based on the only factor that matters in bitcoin for 99.9% of the people: profitability.

[grue]

Sorry, but to say the product is miles away from what it was claimed is patently ridiculous.  Even at the power draw and hashrate I saw, which is IMHO a worst case scenario, the product is somewhat off.  It's not "miles away" by any stretch of the imagination.  20w vs 80w.  Yes, a factor of 4.  So is 1w vs 4w, but that doesn't mean it's a huge amount in absolute terms.

it's nothing in terms of profitability, but it does show that they are incompetent

 Whether it's 20w or 80w, it's still better by a factor of 6 (being generous here, it's more like a factor of 7 or to any GPU solution and at LEAST 3x possibly 4x better than any other FPGA solution... So trying to couch it in terms of "factors" on how off the power estimates were is disingenous.  Put it in terms of absolute values.  It's 80w (and likely the shipping version will be more efficient), compared to 20w.  I'll take the 60w increase in place of my power hungry GPUs screaming in my basement, thanks.  And personally, I'm not going to quibble over 170 - 200 MH/s as a worst case scenario, and likely less than that as well.  Would I like a 1.05 GH/s at 20w?  Hell yes, but the reality and the initial target are anything but "miles away" from each other.  

it's better than GPUs, but that doesn't mean you can overestimate by a factor of 4. It's like saying a hybrid can get 1000 MPG when it can only get 50. is it better than a regular car? yes. but just because it's better, should a manufacturer be able to overstate its claims? no

Fuck seriously.  I am trying to stay out of this, but the rampant hate and complete bullshit some of you are spewing are borderline pathological and far, far over the border of logical or reasonable.  I see mistakes and misunderstandings, but I see little if any outright dishonestly.  But I'll tell you what I do see, is a whole bucketload of a complete lack of understanding or experience in design, business operations, marketing and development.  Anyone who's actually done those things understands exactly what's happened so far and why... anyone who hasn't really shouldn't be speaking.

It's just that you always seem to be biased towards BFL, even when 75%+ of the forums disagree.

[rjk]

It's just that you always seem to be biased towards BFL, even when 75%+ of the forums disagree.

Perhaps because he has seen and tested the product in question, and everyone else has not? I dunno, just a thought.

[CubedRoot]

Stand back gentlemen a wild troll approaches.
How do you get that 75 recent of the forum disagrees with Inaba and BFL.