Butterfly Labs ASICS

[smoothie]

I have just one question: How can you know that your ASICS/rigs you are touting/preordering going to have as much hashing power as you claim on your website unless you have already created the ASICs and benchmarked them.

This question I need answered.

To claim a rig will have a hash rate and not be able to actually sell (more importantly, demonstrate its hashing capabilities), says that you likely don't have the technology at all. Otherwise why not just keep the hardware secret and mine happily for the rest of bitcoinity and make tons more profit?

 

[1715313191]

1715313191

[1715313191]

1715313191

[1715313191]

1715313191

[1715313191]

1715313191

[psilan]

eugh

[Unacceptable]

[OmegaNemesis28]

^ lol
Honestly, they probably won't be as currently advertised judging by the current gen FPGA devices. However, they will be close judging again on that track record.

[Dargo]

It really doesn't matter very much whether BFL ASICs come in below spec or not. Even if they have half the hashing power predicted, they will still crush GPU/FPGA + overwhelm the network. If you have an SC single, you will have 20 Gh instead of the promised 40 Gh. But since difficulty will essentially be determined by all the other BFL ASICs on the network, difficulty will be half what it would have been if BFL had met spec on all the units. So at 20 Gh you will be just as profitable as you would have been at 40. Frankly, I wouldn't mind at all if the units come in under spec, since probably a lot of stupid people would get pissed off and cancel their orders, thus slowing the relative growth of the network, and making my hardware more profitable. So by all means, let's promote this idea that people should get really indignant if the stated specs aren't met.

[OmegaNemesis28]

It really doesn't matter very much whether BFL ASICs come in below spec or not. Even if they have half the hashing power predicted, they will still crush GPU/FPGA + overwhelm the network. If you have an SC single, you will have 20 Gh instead of the promised 40 Gh. But since difficulty will essentially be determined by all the other BFL ASICs on the network, difficulty will be half what it would have been if BFL had met spec on all the units. So at 20 Gh you will be just as profitable as you would have been at 40. Frankly, I wouldn't mind at all if the units come in under spec, since probably a lot of stupid people would get pissed off and cancel their orders, thus slowing the relative growth of the network, and making my hardware more profitable. So by all means, let's promote this idea that people should get really indignant if the stated specs aren't met.

Did you recieve my PM by any chance?

[cuz0882]

It really doesn't matter very much whether BFL ASICs come in below spec or not. Even if they have half the hashing power predicted, they will still crush GPU/FPGA + overwhelm the network. If you have an SC single, you will have 20 Gh instead of the promised 40 Gh. But since difficulty will essentially be determined by all the other BFL ASICs on the network, difficulty will be half what it would have been if BFL had met spec on all the units. So at 20 Gh you will be just as profitable as you would have been at 40. Frankly, I wouldn't mind at all if the units come in under spec, since probably a lot of stupid people would get pissed off and cancel their orders, thus slowing the relative growth of the network, and making my hardware more profitable. So by all means, let's promote this idea that people should get really indignant if the stated specs aren't met.

That's not really true, there will be a decent amount of mining besides ASIC's. A lot people don't have power costs. Then you have other FPGA's and botnets. It will definitely matter less as time goes on.

[OmegaNemesis28]

It really doesn't matter very much whether BFL ASICs come in below spec or not. Even if they have half the hashing power predicted, they will still crush GPU/FPGA + overwhelm the network. If you have an SC single, you will have 20 Gh instead of the promised 40 Gh. But since difficulty will essentially be determined by all the other BFL ASICs on the network, difficulty will be half what it would have been if BFL had met spec on all the units. So at 20 Gh you will be just as profitable as you would have been at 40. Frankly, I wouldn't mind at all if the units come in under spec, since probably a lot of stupid people would get pissed off and cancel their orders, thus slowing the relative growth of the network, and making my hardware more profitable. So by all means, let's promote this idea that people should get really indignant if the stated specs aren't met.

That's not really true, their will be a decent amount of mining besides ASIC's. A lot people don't have power costs. Then you have other FPGA's and botnets. It will definitely matter less as time goes on.

Yes, there will be mining besides ASICS but they will either drop or convert to asics due to the increased difficulty.
Power consumption is now only part of the equation. A $2000 5870 farm still doesnt compare to $600~ Jalapeno one in raw Gh/s. sprinkle in the difficulty and what he says is pretty much true

[Dargo]

It really doesn't matter very much whether BFL ASICs come in below spec or not. Even if they have half the hashing power predicted, they will still crush GPU/FPGA + overwhelm the network. If you have an SC single, you will have 20 Gh instead of the promised 40 Gh. But since difficulty will essentially be determined by all the other BFL ASICs on the network, difficulty will be half what it would have been if BFL had met spec on all the units. So at 20 Gh you will be just as profitable as you would have been at 40. Frankly, I wouldn't mind at all if the units come in under spec, since probably a lot of stupid people would get pissed off and cancel their orders, thus slowing the relative growth of the network, and making my hardware more profitable. So by all means, let's promote this idea that people should get really indignant if the stated specs aren't met.

Did you recieve my PM by any chance?

Yes - reply sent.

In reply to cuz - I agree that non-ASIC mining will continue, and that this will be a factor, but still I think that it will be a small enough part of the network that it really isn't very crucial whether BFL meets spec or not. Missing by half is of course an extreme case. the more likely scenario is that they miss by 20% or so, and this won't make much difference. 

[pieppiep]

Asic design tools probably have some simulation options so you can make an estimate about the speed and energy usage.

About the part about fpga's still being used by people with free electricity, if the difficulty rises a factor 10 and normally the ROI is about a year, it now becames 10 years.
After the difficulty increase the 10 years becames 20 years.
After about 4 years there is 16 years left and another difficulty increase makes it 32.
After 4 more... 28 becames 56...
Unless the price of bitcoin rises enough ofcourse.