Butterfly Labs - Bitforce Single and Mini Rig Box

[kano]

whats the reasoning you guys are using to buy these instead of ICARUS fpga's.

just curious.

Too expensive and too slow.

200 MH/s for $500, really? Thanks, no.

This.

Not "this" - yes he corrected it but it's still not right.

Icarus 380MH/s $569

Not sure where ANY of those other numbers came from ... ... ...

[PulsedMedia]

whats the reasoning you guys are using to buy these instead of ICARUS fpga's.

just curious.

Rig box should get 20Mh/w and was priced (when I bought them) at $25k each. That's $0.50/Mh which is pretty darn good unless you are really bargain hunting for GPUs and the other items needed to run them.

When the subsidy halves, we will see a lot of GPU miner become unprofitable with 2Mh/w to 5Mh/w unless the price rises substantially. I doubt the price will jump like this in lock step with the halving (I've been wrong before) so we should see the same thing we saw when the price went to $2. Miners will quit.

I'm targeting long term goals, so even if price goes down now to 1$ i will keep my miners operational. I see all of this as a risk investment and my goals are set on not what BTC is today, but what it is 1,3,5,10years from now. If best case scenario happens in 10-15years i can retire off my mining earnings from the 2$ days or whatever the low point is going to be. Hence i mostly invest outside money to my rigs. Right now i'm on 2nd investment period and i've yet to sell single BTC (quite the contrary).
My initial miners were bought with 10€ BTC investment from year earlier, which was worth at peak something like 7500€ .... Looking back now i should've sold 90% of my BTC back then XD

[max in montreal]

I have 2 issues I am trying to figure out...

1.they guys making the fpga's can sell them, but for every one they sell, they keep one, and mine with it, eventually drowning us all out.
2.the reasons you give for guys with gpu's dropping out will eventually be the same reasons you guys drop out, your buying the least efficiant fpga.

[malevolent]

Rig box should get 20Mh/w and was priced (when I bought them) at $25k each. That's $0.50/Mh which is pretty darn good unless you are really bargain hunting for GPUs and the other items needed to run them.

How can you tell that rig boxes will get 50GH and use 2.5kW? I wouldn't be surprised if it turned the same as with singles (1.05GH, 19W if I remember right..). Of course I hope this time they will go along a different route, I might buy one myself then (once I raise enough money ofc, $30k, even 25, is a bit steep).

[makomk]

Yes theoretically it would be optimal to always return work as soon as it is found.  Still for efficiency IIRC even GPUs  run an entire "batch" and return all work at the end.  The batch just happens to be small.  Technically the larger batch size may be resulting in slightly higher lost work even in a "normal pool".

I believe GPUs are designed to process batches of computations, so you pretty much have to do it that way. There's no reason why FPGAs can't be designed to return nonces as they're found and continue processing the work though, and unlike with GPUs there's not really an efficiency penalty for decreasing latency this way on the designs I've seen.

There is no "free" lunch when working with a smaller nonce range though.  There is some latency and overhead in setting up work for the processor.

If you're clever you can make this latency more or less irrelevant to the total hashrate on FPGAs, though I don't think the BitForce boards do that either.

[jamesg]

I have 2 issues I am trying to figure out...

1.they guys making the fpga's can sell them, but for every one they sell, they keep one, and mine with it, eventually drowning us all out.
2.the reasons you give for guys with gpu's dropping out will eventually be the same reasons you guys drop out, your buying the least efficiant fpga.

The jump from CPU to GPU along with a huge price spike will be quite hard to replicate IMHO. Bringing on 1Th of FPGA/sASIC equipment will not raise the difficulty significantly and will take quite a while. It will take something like a custom ASCI chip sold in mass to duplicate the jump in difficulty we had from CPU to GPU.

As for buying the least efficient FPGA, I believe it is the jump from GPU to FPGA which matters most to my bottom line. A rig box, as proposed, is so much more efficient than my GPUs that it will pay for it's electricity costs in 1 day of mining whereas it takes my GPUs 15 days to pay for the electricity they use. If I tried to purchase equipment that was twice as efficient as the rig box, it would pay for the electricity in 1/2 day. This is a diminishing return and not the only thing to consider as I would like to make my original investment back.

[stevegee58]

I'll wait for some of the early adopters to get out of mining in a few months and buy a used BFL for 50% off.  

[jamesg]

Rig box should get 20Mh/w and was priced (when I bought them) at $25k each. That's $0.50/Mh which is pretty darn good unless you are really bargain hunting for GPUs and the other items needed to run them.

How can you tell that rig boxes will get 50GH and use 2.5kW? I wouldn't be surprised if it turned the same as with singles (1.05GH, 19W if I remember right..). Of course I hope this time they will go along a different route, I might buy one myself then (once I raise enough money ofc, $30k, even 25, is a bit steep).

BFL has already adjusted their original estimates on the Rig box based on their finding from the single and are quite confident this time around.

[SgtSpike]

whats the reasoning you guys are using to buy these instead of ICARUS fpga's.

just curious.

Rig box should get 20Mh/w and was priced (when I bought them) at $25k each. That's $0.50/Mh which is pretty darn good unless you are really bargain hunting for GPUs and the other items needed to run them.

When the subsidy halves, we will see a lot of GPU miner become unprofitable with 2Mh/w to 5Mh/w unless the price rises substantially. I doubt the price will jump like this in lock step with the halving (I've been wrong before) so we should see the same thing we saw when the price went to $2. Miners will quit.

I'm targeting long term goals, so even if price goes down now to 1$ i will keep my miners operational. I see all of this as a risk investment and my goals are set on not what BTC is today, but what it is 1,3,5,10years from now. If best case scenario happens in 10-15years i can retire off my mining earnings from the 2$ days or whatever the low point is going to be. Hence i mostly invest outside money to my rigs. Right now i'm on 2nd investment period and i've yet to sell single BTC (quite the contrary).
My initial miners were bought with 10€ BTC investment from year earlier, which was worth at peak something like 7500€ .... Looking back now i should've sold 90% of my BTC back then XD

It makes no sense to mine for BTC when you're doing it at a loss.

If you mine 25 BTC for $250 of electricity, but the price per BTC was $5 and you could have bought 50 BTC for $250, then you just lost 25 BTC.

[Inaba]

I have 2 issues I am trying to figure out...

1.they guys making the fpga's can sell them, but for every one they sell, they keep one, and mine with it, eventually drowning us all out.
2.the reasons you give for guys with gpu's dropping out will eventually be the same reasons you guys drop out, your buying the least efficiant fpga.

The jump from CPU to GPU along with a huge price spike will be quite hard to replicate IMHO. Bringing on 1Th of FPGA/sASIC equipment will not raise the difficulty significantly and will take quite a while. It will take something like a custom ASCI chip sold in mass to duplicate the jump in difficulty we had from CPU to GPU.

As for buying the least efficient FPGA, I believe it is the jump from GPU to FPGA which matters most to my bottom line. A rig box, as proposed, is so much more efficient than my GPUs that it will pay for it's electricity costs in 1 day of mining whereas it takes my GPUs 15 days to pay for the electricity they use. If I tried to purchase equipment that was twice as efficient as the rig box, it would pay for the electricity in 1/2 day. This is a diminishing return and not the only thing to consider as I would like to make my original investment back.

That even applies to the BFL Singles.  300+W to 80w is a huge jump, whereas 80w - 20w, not so much.  You can power 80w off a solar panel fairly easily, you can't do that with 300+ watts, which does not even include the cost to dissipate the heat generated (though some climates will make that back up to a degree in the winter time).  

You also need to factor in GPU failure... while we don't have numbers for FPGA failure yet, hopefully it will be more robust in terms of failure to density.  Right now a 5970 is the best bang for the buck, both power and density wise, but they are the highest rate of failure due to craptacular design.  Assuming no major design flaws in the FPGAs, the lack of maintenance requirement and power usage between 80w and 20w is negligible when compared to the same on GPUs.

[PulsedMedia]

whats the reasoning you guys are using to buy these instead of ICARUS fpga's.

just curious.

Rig box should get 20Mh/w and was priced (when I bought them) at $25k each. That's $0.50/Mh which is pretty darn good unless you are really bargain hunting for GPUs and the other items needed to run them.

When the subsidy halves, we will see a lot of GPU miner become unprofitable with 2Mh/w to 5Mh/w unless the price rises substantially. I doubt the price will jump like this in lock step with the halving (I've been wrong before) so we should see the same thing we saw when the price went to $2. Miners will quit.

I'm targeting long term goals, so even if price goes down now to 1$ i will keep my miners operational. I see all of this as a risk investment and my goals are set on not what BTC is today, but what it is 1,3,5,10years from now. If best case scenario happens in 10-15years i can retire off my mining earnings from the 2$ days or whatever the low point is going to be. Hence i mostly invest outside money to my rigs. Right now i'm on 2nd investment period and i've yet to sell single BTC (quite the contrary).
My initial miners were bought with 10€ BTC investment from year earlier, which was worth at peak something like 7500€ .... Looking back now i should've sold 90% of my BTC back then XD

It makes no sense to mine for BTC when you're doing it at a loss.

If you mine 25 BTC for $250 of electricity, but the price per BTC was $5 and you could have bought 50 BTC for $250, then you just lost 25 BTC.

ONLY if you care about today's value - Like i explained, i'm working on assumptions/projections of value after a period of time, not what it is worth today.

EDIT: Actually, at least for me, 7970 is by far the best value. I can get new 7970s for same price as 5970s right now, without the hassle of needing to search and scout, and get out of warranty used cards. No thanks. And i can get them TODAY, means 1½-2months of revenue before BFL Singles would arrive.

[jamesg]

I would never let girlfriend/wife interfere with my business investments, if i see something is going to bring X % ROI annually, all she'd need to know is that it's going to bring the bacon on table.

Lots of american men are pussies when it comes to their women.  The men bring in all the money yet still let their bitches tell them what to do with the money.  Can you believe that?

American women are spoiled cunts.

On another note, my wife knows all about the mining adventure and has approved ever since she saw the first $1k hit the bank account. She even put up with the house being 86 degrees when there was 5Gh running here.

[SgtSpike]

whats the reasoning you guys are using to buy these instead of ICARUS fpga's.

just curious.

Rig box should get 20Mh/w and was priced (when I bought them) at $25k each. That's $0.50/Mh which is pretty darn good unless you are really bargain hunting for GPUs and the other items needed to run them.

When the subsidy halves, we will see a lot of GPU miner become unprofitable with 2Mh/w to 5Mh/w unless the price rises substantially. I doubt the price will jump like this in lock step with the halving (I've been wrong before) so we should see the same thing we saw when the price went to $2. Miners will quit.

I'm targeting long term goals, so even if price goes down now to 1$ i will keep my miners operational. I see all of this as a risk investment and my goals are set on not what BTC is today, but what it is 1,3,5,10years from now. If best case scenario happens in 10-15years i can retire off my mining earnings from the 2$ days or whatever the low point is going to be. Hence i mostly invest outside money to my rigs. Right now i'm on 2nd investment period and i've yet to sell single BTC (quite the contrary).
My initial miners were bought with 10€ BTC investment from year earlier, which was worth at peak something like 7500€ .... Looking back now i should've sold 90% of my BTC back then XD

It makes no sense to mine for BTC when you're doing it at a loss.

If you mine 25 BTC for $250 of electricity, but the price per BTC was $5 and you could have bought 50 BTC for $250, then you just lost 25 BTC.

ONLY if you care about today's value - Like i explained, i'm working on assumptions/projections of value after a period of time, not what it is worth today.

EDIT: Actually, at least for me, 7970 is by far the best value. I can get new 7970s for same price as 5970s right now, without the hassle of needing to search and scout, and get out of warranty used cards. No thanks. And i can get them TODAY, means 1½-2months of revenue before BFL Singles would arrive.

If you mine 25 BTC for $250 of electricity, instead of BUYING 50 BTC for $250, then you are out 25 BTC.

If BTC is worth $1000 in the future, then you'd have $25,000 from mining or $50,000 from buying them directly.

What assumptions/projections are you using that change this?

[PulsedMedia]

If you mine 25 BTC for $250 of electricity, instead of BUYING 50 BTC for $250, then you are out 25 BTC.

If BTC is worth $1000 in the future, then you'd have $25,000 from mining or $50,000 from buying them directly.

What assumptions/projections are you using that change this?

Who said i don't do both?
On another one of my locations electricity is billed annually ...
So if i don't mine today when rates are at $1, but when electricity bill comes and BTC is worth 10$ ... oops
Assuming i actually need to sell BTC to pay that.

Electricity bill always arrives at least month later, so if i mined at 5$ then stop when it's at $1 and then sell BTC to cover electricity while i were mining at 5$ ... ooops!

Only thing what matters is value at sale time - not anything else, absolute nothing else.
If i stop mining while value is low, it's equivalently more time to put off the ROI, 1:1 relation.
Some of us can afford not to dance at the edge of cash flow right now, in order to maximize gain.

Infact, i've yet to sell a single BTC of my mining profits. I used BTC to purchase some of my rigs last summer tho, when BTC was at all time high.

But, oh please, oh please, DO by all means shutdown your rigs down when value visits a little bit lower

[SgtSpike]

If you mine 25 BTC for $250 of electricity, instead of BUYING 50 BTC for $250, then you are out 25 BTC.

If BTC is worth $1000 in the future, then you'd have $25,000 from mining or $50,000 from buying them directly.

What assumptions/projections are you using that change this?

Who said i don't do both?
On another one of my locations electricity is billed annually ...
So if i don't mine today when rates are at $1, but when electricity bill comes and BTC is worth 10$ ... oops
Assuming i actually need to sell BTC to pay that.

Electricity bill always arrives at least month later, so if i mined at 5$ then stop when it's at $1 and then sell BTC to cover electricity while i were mining at 5$ ... ooops!

Only thing what matters is value at sale time - not anything else, absolute nothing else.
If i stop mining while value is low, it's equivalently more time to put off the ROI, 1:1 relation.
Some of us can afford not to dance at the edge of cash flow right now, in order to maximize gain.

Infact, i've yet to sell a single BTC of my mining profits. I used BTC to purchase some of my rigs last summer tho, when BTC was at all time high.

But, oh please, oh please, DO by all means shutdown your rigs down when value visits a little bit lower

Ahhhh, it all makes sense now!  Annual billing especially.  It's like some sort of weird hedge on future prices.

I was kind of wondering if you might be talking about the effective 30-60 day lag between monthly billings of electricity, but didn't even know annual billing existed.

Sorry, I've just dealt with too many idiots here who didn't understand it.  I assumed you weren't any different.  My bad. 

[triplehelix]

I would never let girlfriend/wife interfere with my business investments, if i see something is going to bring X % ROI annually, all she'd need to know is that it's going to bring the bacon on table.

Lots of american men are pussies when it comes to their women.  The men bring in all the money yet still let their bitches tell them what to do with the money.  Can you believe that?

American women are spoiled cunts.

On another note, my wife knows all about the mining adventure and has approved ever since she saw the first $1k hit the bank account. She even put up with the house being 86 degrees when there was 5Gh running here.

on that same note, i was giving my wife a quick overview of bitcoin, and told her i was going to be investing in it.  she still had the glazed over look in her eyes from the bitcoin explanation, but she didn't miss a beat with the, "i trust you to do whatever it is you think you should do."

[jamesg]

I would never let girlfriend/wife interfere with my business investments, if i see something is going to bring X % ROI annually, all she'd need to know is that it's going to bring the bacon on table.

Lots of american men are pussies when it comes to their women.  The men bring in all the money yet still let their bitches tell them what to do with the money.  Can you believe that?

American women are spoiled cunts.

On another note, my wife knows all about the mining adventure and has approved ever since she saw the first $1k hit the bank account. She even put up with the house being 86 degrees when there was 5Gh running here.

on that same note, i was giving my wife a quick overview of bitcoin, and told her i was going to be investing in it.  she still had the glazed over look in her eyes from the bitcoin explanation, but she didn't miss a beat with the, "i trust you to do whatever it is you think you should do."

Good woman you got there. Mine did the same until the bank account bloated up a bit.

[PulsedMedia]

Ahhhh, it all makes sense now!  Annual billing especially.  It's like some sort of weird hedge on future prices.

I was kind of wondering if you might be talking about the effective 30-60 day lag between monthly billings of electricity, but didn't even know annual billing existed.

Sorry, I've just dealt with too many idiots here who didn't understand it.  I assumed you weren't any different.  My bad. 

Same effect happens when you simply can pay off your electricity bills regularly without selling any BTC.
It's nothing special, just a matter of financial effectiveness.

[WhitePhantom]

I would never let girlfriend/wife interfere with my business investments, if i see something is going to bring X % ROI annually, all she'd need to know is that it's going to bring the bacon on table.

Lots of american men are pussies when it comes to their women.  The men bring in all the money yet still let their bitches tell them what to do with the money.  Can you believe that?

American women are spoiled cunts.

On another note, my wife knows all about the mining adventure and has approved ever since she saw the first $1k hit the bank account. She even put up with the house being 86 degrees when there was 5Gh running here.

on that same note, i was giving my wife a quick overview of bitcoin, and told her i was going to be investing in it.  she still had the glazed over look in her eyes from the bitcoin explanation, but she didn't miss a beat with the, "i trust you to do whatever it is you think you should do."

My experience was similar when I told my wife I thought we should borrow $30k for a rig box.  I look forward to the "bank account bloating" part that gigavps mentioned. :-D

[PulsedMedia]

Well, i bited and purchased my first 2 units.
Money reserved for next 5 after these arrive.

and i still got 7 new GPUs to setup too :/