Topic: What else can you do with an Asic USB Miner? - page 2. (Read 16121 times)

well make sure you get one for $5. They don't ROI at higher prices if you are interested in profiting.

I'm thinking about getting one to use with an RPi, think it would be kind of neat, but probably never pay itself off.

Well I have a Raspberry Pi already + $10 Amazon gift card.... Hmmm...

They're not designed to produce numbers quickly - they're designed to produce answers (valid shares) with a minimum of bandwidth involved.  Remember that people are hanging dozens of these ASICs off heavily chained USB hubs, and there's no noticeable performance loss.  The per-device bandwidth is insanely low, by design.

I'm still open to any ideas, though.   It's just that the guys who did the crypto on bitcoin were actually damned good at what they did, and forked off into their own little corner to ensure that it was totally separate from anything else that matters.  It's quite impressive, actually.

You are right, 2 bits per second isn't enough, i thought it would have been faster. in which case you are right, these couldn't be used for that. I thought that at its easiest difficulty it would produce numbers faster.

The difficulty is not adjustable - the miners simply return everything that has a difficulty of 1 or greater (which, conveniently, means everything with one of the end 32-bit output registers all zero) and let the host sort it out.

If you're doing what I think you're suggesting (using the calculated nonces as a pseudorandom stream), you'd be getting something measured in bits per second - a 330HM miner should find a difficulty 1 block roughly every (2^32/330M) = ~13 seconds.  So a 32-bit nonce per 13 seconds.  This isn't particularly useful as a stream cipher - 2 bits per second isn't stellar.

You've been insisting that there is another obvious use for these, "think harder," etc - but from my understanding of how the ASICs work, they are really only good at performing the bitcoin mining calculation or something else deliberately fit into that framework.

I'm not aware of any other type of cryptographic operation that looks anything like what bitcoin does - the people who designed it have done a very nice job of sticking it over in its own little corner of the crypto space and ensuring that it's only useful for bitcoin.

I'd love to be shown wrong, and I'm trying to convince you to do so.  I'm just not entirely convinced you have a solid grasp of the exact nature of the calculations performed and optimizations used by the hardware to do what they need to do, as fast as possible.

perhaps create your own local blockchain with low difficulty, set up a few USBees and have a very elaborate 'hash race' betting track

Difficulty for block hashing algorithm is totally meaningless and has nothing to do with general purpose sha256 applications. ASICs in their current form, and likely forever more when designed for bitcoin, cannot do general purpose sha256.

What would happen if you give it difficulty "0" Even with a Difficulty of "1" i suspect that it would return enough bits to do, say secure voice over IP? Use one for encrypt and one for decrypt in a fulll duplex setup.

I maintain the Cryptohaze password cracking suite, I'm a panel member on the Password Hashing Competition, and I've spoken at several conferences on password cracking and encryption.  I know my way around encryption.

https://en.bitcoin.it/wiki/Block_hashing_algorithm

Per the best of my understanding, the bitcoin ASICs take the partial result of the first inner SHA256 operation (the first 64 bytes), and then calculate another SHA256 block starting with the partial state (the 8 32-bit registers) and the 16 remaining bytes, incrementing the nonce each round.  They perform the outer SHA256 on the result, and look for a result matching difficulty 1 or greater.  If this is found, they return the nonce used to find it else say they didn't find it.

You appear to be claiming that the ASICs are a general SHA256 engine that lets you put arbitrary data in and get the SHA256 hash out - or at least this is what it sounds like to me.

I'm asking, "What other application do you have that fits with the manner in which the ASICs actually operate?"  I'm not aware of one.

You are kind of asking me to teach you about encryption and how to do it. Regardless, of what they were made for, for a specific input string, you get a specific psuedorandom output string. Without knowing the input, you can't predict the output. That is all that is needed for encryption/decryption, the rest is software. Considering how fast these guys are, they would lend themselves to real time applications.

I would think they'd have started a few fires by now if that were the case.

sew a few into a blanket for the winter!!

Yes it is. I will be facing it very soon with my erupter blades 

can anyone confirm that a hot USB Asic Miner can light a cigarette?

Quite the conundrum isn't it?

Why not turn it to an alt coin like ppcoin or zetacoin. The likelihood of you finding a block is very low. You may go years without finding one on a block erupter. Mining an alt coin atleast gives you something and you can exchange it for btc whenever you want. But should the price of the alt coin rise you could potentially make a little pocket cash 

Shhh! Haha, I meant it more of a 'I'll just let it sit and do it's thing'. Much like finding a $50 in your winter jacket the first time you use it from last season.

You keep going on about this.

Could you explain how to use a bitcoin ASIC for encryption/decryption?  They take a data block, iterate a nonce until a target is found, and return the nonce.  At least as far as my understanding of them goes.  They're not a "general SHA256 engine" on a USB stick - they are a very, very specific bitcoin miner.

Basically, these could be used in any application that requires enryption and decryption. A low cost solution with the right software.

Maybe you can use the hashing power to accelerate a software.