BFL releases actual specs. ACTUAL SPECS!

[enmaku]

http://codinginmysleep.com/butterfly-labs-releases-more-data/

[juhakall]

Can't really blame them if they deliver more than was promised. It just seems weird that the Jalapeno only gets a ~29% increase, while the Single and Minirig both get 50%.

[DeaDTerra]

Can't really blame them if they deliver more than was promised. It just seems weird that the Jalapeno only gets a ~29% increase, while the Single and Minirig both get 50%.

This is in fact what happens in almost all chip production.
Take for example 6950 and 6970 they are the same chip yet they perform differently this is due to that the better performing chips becomes the 6970 while the worse performing becomes the 6950.
Probably the same is true for the BFLS, maybe the good chips becomes Singles and minirigs while the worse performing becomes singles.
Just a guess.
//DeaDTerra

[Jack1Rip1BurnIt]

Can't really blame them if they deliver more than was promised. It just seems weird that the Jalapeno only gets a ~29% increase, while the Single and Minirig both get 50%.

Well, from what I can gather from the article it all has to do with the power. If you notice the trend is 1 gh/s per watt. Therefore the total max watts on the jalepeno is 4.5 because that is the maximum output for dual usb 2.0 outputs. Crazy hashrates/watts either way...very interesting futureahead of us.

[BCMan]

 1Gh per watt? Yeah, right.

[enmaku]

Can't really blame them if they deliver more than was promised. It just seems weird that the Jalapeno only gets a ~29% increase, while the Single and Minirig both get 50%.

Power consumption, friend, power consumption.

Say before they were thinking their numbers would let them pull off < 2.5 watts, that would let them run off a single USB port without having to worry about whether your port is a 500ma or 1000ma port, without going to the double-plug USB cables or without having to think about USB 3.0 compatibility. With the upgrade they threw the old 2.5 watt max out the window and went to the most you can possibly pull from a USB 2.0 port if it's new or if you have a double cable, 5 watts. I'm guessing they could probably clock it way higher if the power and heating were available, but they're limited by USB power since they don't want to go to wall adapters.

[enmaku]

1Gh per watt? Yeah, right.

[Jack1Rip1BurnIt]

If these performance claims are true then the bitcoin network will be unbelievably secure and we would more than likely see thousands of terahashes before the 12.5 reward half.

[dust]

The SC single board rendering that BFL released had 8 sha256 chips.  I assume the jalapeno will have 1.  It it even more strange now that the single gets more hash rate out of each chip.  The single is now doing 7.5Gh/s/chip compared to the jalapeno's 4.5Gh/s/chip, but they are equally energy efficient.

This announcement is probably in response to bASIC's announcement.  It doesn't increase my confidence in BFL as if they were scamming it would also make sense to make this kind of announcement.  Pictures of the actual products and working demos would be nice...

[enmaku]

The SC single board rendering that BFL released had 8 sha256 chips.  I assume the jalapeno will have 1.  It it even more strange now that the single gets more hash rate out of each chip.  The single is now doing 7.5Gh/s/chip compared to the jalapeno's 4.5Gh/s/chip, but they are equally energy efficient.

This announcement is probably in response to bASIC's announcement.  It doesn't increase my confidence in BFL as if they were scamming it would also make sense to make this kind of announcement.  Pictures of the actual products and working demos would be nice...

I don't have pics or working demos for you, but I can tell you why the jally is only doing 4.5gh/s while the single is getting 7.5gh/s per chip - they didn't want to add a power brick. This limits them to the 5 watts they can reasonably expect to pull from a USB port (either a 1000ma port or using a dual USB adapter with 500ma ports) Assuming they had 7.5 watts to work with I'd bet the jally could pull 7.5 GH/s but it would then require a power adapter.

[Monkey1]

Still no idea of when they are likely to ship and in one of the posts, Josh states that they dont know how many of each type they will be able to ship initially.  Surely with 4 weeks to go to their (revised) stated shipping date, they should know a little more that that?

[Jutarul]

1Gh per watt? Yeah, right.

Why does 1J/GH trigger my bullshit detector?

[Luno]

I don't think the specs are based on actual measurements. A 1500W Power supply would have higher efficiency than the 4.5W one. The figures are estimates based on board design not from the wall measurements on production / prototype -boards. Their headline also don't claim the specs to be measured.

[Monkey1]

With 4 weeks to go you would expect them to have some actual measurements surely? 

[Shadow383]

With 4 weeks to go you would expect them to have some actual measurements surely? 

This announcement is probably in response to bASIC's announcement.  It doesn't increase my confidence in BFL as if they were scamming it would also make sense to make this kind of announcement.  Pictures of the actual products and working demos would be nice...

Noooo! Don't forget what Inaba said! You guys are being entitled and unreasonable.

Look! They've used your millions of dollars and made an extruded cube with rounded edges and applied their logo to it in Solidworks.

[Fcx35x10]

BFL had to adapt to competitions 

[crazyates]

A 1500W Power supply would have higher efficiency than the 4.5W one.

Umm where was it ever said that the "4.5W one" (I'm assuming you're talking about the Jalapeno) would even HAVE a power supply? The Jalapeno receives it's power from a DC feed through the USB. Its efficiency would be whatever your computer's PSU's efficiency is.

[VeeMiner]

these numbers sound great, I hope BFL can deliver...

[Fcx35x10]

these numbers sound great, I hope BFL can deliver...

+1

I agree

[Monkey1]

But When?