Topic: 1GH/s, 20w, $500 — Butterflylabs, is it a scam? - page 27. (Read 123109 times)

Nice.

Not implying this is the case AT ALL, but wouldn't it be epic if it turned out that the mining circuit was acausal logic implemented by arranging the gates in a closed time like curve that allow the system to only attempt solutions which are valid and thus actually achieving infinite speed. ... but then Butterfly labs couldn't productize that because it would make bitcoin worthless, so instead they rate-limit it to realistic speeds in order to maximize their sales. Again, I'm not even suggesting this is the case, I just let my mind wander.

Not implying this is the case AT ALL, but wouldn't it be an epic scam if Inaba gave a down-to-earth review proving that the magic box works as described, several hundred orders roll in at full price, then it turns out he was in cahoots.....  Again, I'm not even suggesting this is the case, I just let my mind wander.

Have you signed an NDA with them?  Have you met them in person yet?

We plan on taking at least some video.  They gave me approval to disassemble it and take pictures (no taking the heatsinks or stuff off, which is understandable - don't want to damage anything.) - so I will be doing that.  It may come down to me doing a physical review of it tomorrow and a live test on Saturday as my schedule is crammed with the holiday, so if the live test slips a day, I would not consider that defaulting on any bets or what have you.

I hope Inaba will make some youtube videos showing the system in operation. That would be great.

I predict some epic crow-eating in 3...2...1...

Forget these Butterfly lab guys, this one actually resembles a butterfly.

Is it easy to cascade a few boards together?  Or how many boards can you install on a PC?

Sweet! Will you be using the unit with a Windows or Linux machine, ie. is it the Windows or Linux driver's source code you've been given?

The software will support Linux, Windows and I believe Mac. 

I am reviewing the driver source code prior to compiling it for tomorrows test as we speak.  I don't know how much or little BFL wants me to talk on the subject at the moment, so I am not providing any additional information.  But to the best of my knowledge at this point, the live demo is still on tomorrow evening CST at my data center.  They've already OK'd me to keep a unit overnight for further testing at another location.  I will inquire as to if they mind if I disassemble the unit and take pictures as well, presuming the tests go as BFL expects tomorrow.

Well I'll have one...

If it works as advertised then it's bang on the price point where I can sell all my energy-guzzling GPUs to eXtReMe GaMeRz, replace with these boards and even end up with a higher overall hashing rate.

In England, plain mining with even highly-optimised GPUs (i.e. converting all mined BTC directly into local currency, in this case GBP) isn't profitable in anywhere near enough amounts to justify the cost and time spent messing with the kit.

I've accounted for some 'loss' by the non-financial benefits (a bit of fun to build the stuff, support of a non-governmentally-controlled cryptocurrency, etc.), but also don't immediately cash my mined BTC out - I use them to buy a different commodity which I'm profiting on, making up for the loss.

My electricity bill will be huge but my gas bill will be tiny - I haven't had to use the gas central heating since I've had the BTC miners running

The big difference with replacing my entire hashing power with these boards is the very low power consumption, and therefore the low heat dissipation. This would allow me to either store the rig in my other house, or turn it into an art object (industrial design-style) by using my old Apple G4 Cube as the controller, and bunging all the FPGA boards into a similarly-designed acrylic cube

Hopefully it's not a scam. I'll be trying a board from the other vendors too, but this particular unit has higher hash rates at a lower price, so worth trying out first.

There is little information about the software platform that the 'control code' runs on, since a host computer is required to manage the embedded hardware boxes. I'll be working on this - I'll only be cancelling my order if the only option is a pre-compiled Windows binary, because I don't use Windows, and having source to compile from will make me more confident that there's nothing dodgy in the 'control code'.


Personally, I'm surprised that the website still claims that the 'early-adopter' versions (the first 100 built) are still available for sale. Regardless of the nay-sayers on this thread, it's only *100* relatively inexpensive boxes. I'd have thought that they'd be all sold by now. Perhaps they already have, and the website simply hasn't been updated... and I won't receive my unit for months and months...

We will see. If the lead-time is extreme then I'll try a ztex board for testing...

Hi wormbog,

I am one of these non-American readers, but I will not go back to work. I even wasn't before reading... 
I prefer keep lying in my hammock and wait for the demonstration and proof (or not)...

Have fun...

Sunny regards from the paradise,
 Eckmar

Happy Thanksgiving, Everyone!

Tomorrow's the big day when Butterfly Labs (or is it BF Labs?) provides a public demonstration of their revolutionary new hashing technology. Either it will turn out to be real, which would be awesome, or it will be delayed or disproven, which will give many of us on this board smug satisfaction. Either way we'll all have something to be thankful for!

In the mean time, enjoy your turkey and stuffing! For those non-American readers out there, get back to work!

That is kinda the point.
As an example:
SHA-256 is used in IPSEC for example however 1 bitcoin GH =2 billions hashes per second.  @ 512KB per block that is =2tbps (2 terrabits per second).

So nobody making SHA-256 accelerators is going to make one that fast.  An high end IPSEC accelerator might have 1Gbps throughput = 1Bitcoin MH/s.

Same thing w/ accelerators used to offload website login functionality.  Even an accelerator designs to allow 1 million simultaneous logins would only result in 1 Bitcoin KH/s.  Bitcoins usage profile is so off the charts compared to anything else that nobody makes off the shelf chips designs to work on the scale required for Bitcon simply because no market exists for them.

Except for thw whole "doesn't make sense at all because no one makes a terabit/sec sha256 chip" part, right?

There are plenty of SHA accelerators, I just havent seen anything in this performance range, or even close to it. That doesnt mean they dont exist, if nothing else, I kind of doubt the NSA uses FPGAs; Id just like to know what shelve to look at

IMHO I already asked about this and somebody ( I think D&T ) told me that they never make a hardware SHA256 thingy because there is no demand for it etc.

Holding my breath until 25 Nov. If they delay yet again I will shout "SCAM !!!" and if not "BUY BUY BUY !!!" LOL

What off the shelve sha256 unit would that be?

Both sides obviously have strong opinions. Money talks. 150 BTC.