What's the expected probabilistic error rate? I'm seeing approximately 1% HW errors (apparently false nonces) reported.

The Satechi hub is also specified as only being able to drive 250mA per port REGARDLESS of current supplied, so it may not even work with one BE USB.

I've been having a bear of a time with btct.co timing out on me today.  Is something up with the site?

A number of pools implement variable-difficulty shares so that any given worker is submitting at a roughly constant rate whether it's a CPU-mining Raspberry Pi or an overclocked 6-GPU monster.  It helps the pool by preventing it from being share-spammed into oblivion by large rigs, and helps the miner by giving relatively stable error bounds on share submission.

Ka-ching!   My own shares are outperforming my mining this week, then.

Serious questions about these devices' implementation of Icarus protocol though:

(1) are they interruptible (sending down a new work unit discards the one in progress immediately) to minimize stales?  This is a headache with my BFL FPGAs.
(2) If not interrupted, will they scan the entire noncerange or abort after hitting a share?
(3) Do they report found shares immediately or at the end of the work?
(4) Do they implement notification on exhausting the noncerange without finding a share?

I ask this because I'm considering rolling a little custom miner app just for them.

Mind if I try to hop on before this thing departs? I'll take 5 myself.

burnside's not gone, just preoccupied.  Between LitecoinGlobal, btct.co, Kättare Hosting, LTC-MINING, ASICMINER-PT, and whatever else he's gotten into since I last looked, administering the pool is probably just not a high priority.  I got an email from him today regarding ASICMINER-PT so I know he's still lurking.

Thanks!


Five... four... three...

Lovely.  Now to find another block and see if everything works as expected

Think I'll hop in the pool and give it a try.

Cannonball!!!! *sploosh*

StrongCoin is a hosted wallet service that deducts 1% fee from every spend, 0.5% to the StrongCoin fee-drop and 0.5% to the miner that confirms the transaction.  They're apparently a fairly well built hosted wallet in that private keys are only ever decrypted in the client's browser, but also have the atypical habit of returning change to one of the original source addresses rather than a new address.

Given that I've been blown off the leader board with the better part of a day to go, it wouldn't surprise me.  I think there may well be some collector's value to this particular batch of ten devices as the first hardware shipped by ASICMINER.

Blade fuse, blade board... needs an old-fashioned knife switch, to keep up with the "cutting edge" theme. 

Okay, I'll stop with the hurricane of puns.

2 @ 61.  I have to hedge my bets, since the butterflies haven't left the chrysalis.

You accidentally 1000 chips?  (Ba-dum-tish!)

Sorry, couldn't resist.   

It looks like typical "geek" small business to me: questionable customer service, a lot of margin for error in delivery dates and effectiveness, but they do eventually ship product and it fundamentally works... took them about nine weeks to get me an FPGA Single but it's whirring away on my desk like a ... little cube of hashing machinery. 

Yes, that's one way to get some coins. 

Or did you mean something more like this:

http://fc05.deviantart.net/fs5/i/2005/013/1/8/Seagull_by_kaykaykit.jpg

I don't know that it's wasted power exactly.  The way I see it, there's something like a domino effect going on: FPGAs are uprooting GPUs for the sha256d proof of work, while GPUs have just made their debut on the scrypt proof of work.  As for whether CPU-mining LTC is worth the trouble: look at the marginal power cost if the machine was going to be on anyway, or at the total power cost if not.  Should they ever hit the market, ASICs might well clobber current FPGA board owners, though since the scrypt proof of work requires much more memory, those board owners might be SOL (barring the Butterfly Labs metamorphosis trade-up offer).

An Atom D525 has a marginal power cost of only about 3W over idle for 5.5 kH/s, which (while small) is a net win.  Extrapolating current values, that's an expected payback of 18.62 LTC a month, which should cover even fairly outrageous energy pricing if the machine is otherwise left on anyway.  On the other hand, that same desktop with SSD and nVidia ION graphics has an active-idle draw of about 20W at the wall; this makes it a net loss even at 0.075 USD/kWh if the machine is not required to be left on.  (You can squeeze ~1 MH/s out of the ION mining BTC, but that's not enough to change the equation.)

Then there's always the gaming potential of a good GPU.  Let it mine for you when you're not keeping it busy blowing the living daylights out of something. 

I've been a fairly long time lurker myself and find the "newbie jail" to be an annoyance, but given the way Bitcoin attracts troublemakers, I can kind of understand it.