Butterfly Labs - Bitforce Single and Mini Rig Box

[SgtSpike]

Really, the BTC mined during testing amounts to a drop in the bucket compared to what they are selling these things for.  No, I do not think they are purposefully keeping them on hand longer than necessary for testing purposes.  They want to get them out the door so they can get more orders in.

It wouldn't really make sense for BFL not to test them by mining for their own profit. I don't believe that rotating through stock, say keeping 30 singles running at a time would be negligible.

30*(832 Mhash/s) = 24.96 Ghash/s

At current difficulty (1498294) with an exchange of $4.7 per BTC (about what it is at time of posting), that would be a little over $550 per a week for them.

Seems like a decent amount of money while further ensuring that the product works.

Well, think about it from a different perspective.  If they only sold 30, and they tested them for 12 hours straight, they'd have 25GH/s for 12 hours, and make about $39 worth of BTC.

But they sold those 30 units for $17,970

Do you see now why I said that the mining income is a drop in the bucket?

Even if they kept them for a week, they'd only make $550, as you said, which is less than the price of ONE unit.  Given that they'll get more orders the quicker they are sending the units out, I can't imagine they would hold back for a week on 30 of them, just to make enough mining profit to equate to one more sale of the units, while potentially losing out on many more sales of units because of the delay.

Make sense?

[fcmatt]

Really, the BTC mined during testing amounts to a drop in the bucket compared to what they are selling these things for.  No, I do not think they are purposefully keeping them on hand longer than necessary for testing purposes.  They want to get them out the door so they can get more orders in.

It wouldn't really make sense for BFL not to test them by mining for their own profit. I don't believe that rotating through stock, say keeping 30 singles running at a time would be negligible.

30*(832 Mhash/s) = 24.96 Ghash/s

At current difficulty (1498294) with an exchange of $4.7 per BTC (about what it is at time of posting), that would be a little over $550 per a week for them.

Seems like a decent amount of money while further ensuring that the product works.

Well, think about it from a different perspective.  If they only sold 30, and they tested them for 12 hours straight, they'd have 25GH/s for 12 hours, and make about $39 worth of BTC.

But they sold those 30 units for $17,970

Do you see now why I said that the mining income is a drop in the bucket?

Even if they kept them for a week, they'd only make $550, as you said, which is less than the price of ONE unit.  Given that they'll get more orders the quicker they are sending the units out, I can't imagine they would hold back for a week on 30 of them, just to make enough mining profit to equate to one more sale of the units, while potentially losing out on many more sales of units because of the delay.

Make sense?

If I was them I would sell a unit and make just enough to get myself a free unit to leave on a shelf to mine.
So if they cost me 225 to make and sell I would break even.

So if I sold 100 units I would be mining with 100 units. This all assumes this is pretty much a 1 person gig going
on and he does not have a lot of overhead.

But that is just me. Eventually the market will be saturated for these and business will trickle off unless you can
offer much more bang for the buck. This way you have quite a few "free" units making  you continuous money long term.
Once you are out of electricity and cooling.. then just sell all the new ones that come in unless you can find more space
to put them.

[DeathAndTaxes]

Sgt. Spike is right.  If mining had a high enough ROI% that they would want to delay orders a week or a month then it would be best to delay orders forever and simply keep the units to mine privately.

[Syke]

But they sold those 30 units for $17,970

A wise man once said, "profit matters and revenue means nothing". So how much of that is profit? 20% would be $3594.

Even if they kept them for a week, they'd only make $550

And if they kept them for 4-6 weeks, they'd make an additional $3300 in profit. They just doubled their profit. Who doesn't like to double their profits?

[DeathAndTaxes]

But they sold those 30 units for $17,970

A wise man once said, "profit matters and revenue means nothing". So how much of that is profit? 20% would be $3594.

Even if they kept them for a week, they'd only make $550

And if they kept them for 4-6 weeks, they'd make an additional $3300 in profit. They just doubled their profit. Who doesn't like to double their profits?

If they kept them for 40-60 weeks they'd make $33,000.
If they kept them for 400 to 600 weeks they'd make $330,000.
If they kept them for 4,000 to 6,000 weeks they'd make $3.3 million

So why even sell them?

Money for nothing and you chicks for free.

[stevegee58]

IF they kept them for 40-60 weeks they'd make an additional $33,000 in profit.  So why even sell them?

Oh I dunno.  Because they don't last that long?

These arguments are silly.  This is like people selling trading systems.  Everyone criticizes them saying, "Well if your system were really profitable you wouldn't be selling it."

[RoadStress]

Scanned a few pages and didn't see anything about the Rev3.  If you order today are you guaranteed a Rev3 single with the heat pipe?

Thanks!

PS - Sent a note to Sonny.  Hoping this doesn't take 2+ weeks to answer like last time 

Waiting for an answer too!

[this time]

They're making these for less than 300. The spartan double boards cost about $370? less? to make and these are previous generation chips.

Most likely the initial power estimates were based on 45nm and sometime during the process they got a line on some cheaper chips and thus the giant leap in power consumption. When they run out, the single ver. 2 will be announced, for delivery in 4-6 weeks.

[Energizer]

But they sold those 30 units for $17,970

A wise man once said, "profit matters and revenue means nothing". So how much of that is profit? 20% would be $3594.

Even if they kept them for a week, they'd only make $550

And if they kept them for 4-6 weeks, they'd make an additional $3300 in profit. They just doubled their profit. Who doesn't like to double their profits?

+1

[rjk]

They're making these for less than 300. The spartan double boards cost about $370? less? to make and these are previous generation chips.

Most likely the initial power estimates were based on 45nm and sometime during the process they got a line on some cheaper chips and thus the giant leap in power consumption. When they run out, the single ver. 2 will be announced, for delivery in 4-6 weeks.

You have some proof of your numbers? I'm sure everyone here would love to see some hard facts. Lets start with the chips themselves, perhaps you know exactly what they are and how much they cost? No? Then stfu.

[this time]

They're making these for less than 300. The spartan double boards cost about $370? less? to make and these are previous generation chips.

Most likely the initial power estimates were based on 45nm and sometime during the process they got a line on some cheaper chips and thus the giant leap in power consumption. When they run out, the single ver. 2 will be announced, for delivery in 4-6 weeks.

You have some proof of your numbers? I'm sure everyone here would love to see some hard facts. Lets start with the chips themselves, perhaps you know exactly what they are and how much they cost? No? Then stfu.

Spartan 6 chip = 150-160 confirmed. Total double board cost =$370? reasonable estimate.

BFL selecting something older indicates they got a better deal.

Estimated power consumption going from 20W to 85W indicates there was a change in plan.

single ver.2 will be the elemental part of the mini rig box/rig box that is estimated to achieve similar mh/w to 45nm chips and is being discussed now by bfl themselves.

[rjk]

They're making these for less than 300. The spartan double boards cost about $370? less? to make and these are previous generation chips.

Most likely the initial power estimates were based on 45nm and sometime during the process they got a line on some cheaper chips and thus the giant leap in power consumption. When they run out, the single ver. 2 will be announced, for delivery in 4-6 weeks.

You have some proof of your numbers? I'm sure everyone here would love to see some hard facts. Lets start with the chips themselves, perhaps you know exactly what they are and how much they cost? No? Then stfu.

Spartan 6 chip = 150-160 confirmed. Total double board cost =$370? reasonable estimate.

BFL selecting something older indicates they got a better deal.

Estimated power consumption going from 20W to 85W indicates there was a change in plan.

single ver.2 will be the elemental part of the mini rig box/rig box that is estimated to achieve similar mh/w to 45nm chips and is being discussed now by bfl themselves.

eldentyrell has all but debunked the assumption of a Spartan 6, and has posted a 10 BTC bounty in his thread for the output of the JTAG pins. I'd trust his judgement first, but I'd also like to see someone get that info and give it to him. I don't have the knowledge of how to do so however.

[JWU42]

Received the following response about 3 hours ago to my query about Rev 3 single and Rig Box

Yes, orders today will receive Rev 3 and lead time is still 4-6 weeks and Rig box lead is 12-15 weeks.

Thanks for your interest.

Sonny K.
BFLabs, Inc.

[this time]

They're making these for less than 300. The spartan double boards cost about $370? less? to make and these are previous generation chips.

Most likely the initial power estimates were based on 45nm and sometime during the process they got a line on some cheaper chips and thus the giant leap in power consumption. When they run out, the single ver. 2 will be announced, for delivery in 4-6 weeks.

You have some proof of your numbers? I'm sure everyone here would love to see some hard facts. Lets start with the chips themselves, perhaps you know exactly what they are and how much they cost? No? Then stfu.

Spartan 6 chip = 150-160 confirmed. Total double board cost =$370? reasonable estimate.

BFL selecting something older indicates they got a better deal.

Estimated power consumption going from 20W to 85W indicates there was a change in plan.

single ver.2 will be the elemental part of the mini rig box/rig box that is estimated to achieve similar mh/w to 45nm chips and is being discussed now by bfl themselves.

eldentyrell has all but debunked the assumption of a Spartan 6, and has posted a 10 BTC bounty in his thread for the output of the JTAG pins. I'd trust his judgement first, but I'd also like to see someone get that info and give it to him. I don't have the knowledge of how to do so however.

I'm not saying they used spartan 6. They did not. I'm using it as a comparison. ie If people can make spartan 6 boards and sell them at approx. 50+% higher and that stands up as the market price with multiple sources, then if bfl are using older chips they are doing it because they can achieve a better margin which would put their cost well under $400. My guess was 300 because it was seemingly enough difference to make them totally change their plans and redesign.

[DeathAndTaxes]

So you added the cost of 2 FPGA and assumed that is what the final product costs?

No PCB required, no USB controller, no DC to DC power supply, no production costs, no testing costs, no yield issues.

Just drop two FPGAs in a baggie and ship it out by first class mail with a bunch of stamps.  Who knew FPGA miners were that easy to make.

[bombo999]

[bombo999]

[rjk]

http://www.butterflylabs.com/wp-content/uploads/2012/03/easyminer-2.png

http://www.butterflylabs.com/wp-content/uploads/2012/01/easyminer-1.png

Nice find. Definitely obvious that they are targeting the absolute noobs. That application when released will blow all other FPGAs out of the equation. Hopefully it will support other pools later, although EMC is of course an excellent choice, and Inaba is right next door to assist with compatibility issues etc.

[DiabloD3]

Wait, hold up, does that say Paypal?

[rjk]

Wait, hold up, does that say Paypal?

EMC offers a PayPal cashout option, and they must have made it work with the API.