$10 per GH/s at this point has no chance of ever even coming within a ballpark of breaking even.
There are preorder options a low lower than that and it remain to be seen if they will break even.

As for renting?  Nobody is going to rent you a rig for less than what it can make.  If you owned a machine which printed an authentic $100 bill each day would you rent it to me for $50 a day?

So your options are
a) get ripped off "renting" hashing power at 400% over what it can possibly make.  There are plenty of places that do that cex, cloudhashing.
b) pre-order and hope your rig arrives and you can beat difficulty, sooner and cheaper is better than later and more expensive.  It is a risk and there will be winners and losers.
c) wait and accept that the risk will be lower in the future but the potential rewards will also be lower. 

Nice summary however I believe KNC is using a 4 die concept.  Their error reports refers to die 0,1,2,3.  Google "KNC die zero error".  The large size of the package (45mm x 45mm) also would indicate that.  It is unlikely they are using a giant package for a tiny die, and equally unlikely they have a massive single die underneath that package lid.  It will be interesting the first time a KNC module dies out of warranty.  Dissolving the package would give some good insight.

Mining was my first love (I have over 2,000 posts in the mining sub forums going back over two years).  My 40 GPU farm (watercooled I might add because I am crazy) gave me the revenue to start a Bitcoin related company. The waning of GPU mining and the demands of TC, LLC left me with little time for mining.  For the past couple months I have had too much spare time (banks, lawyers, governmental nonsense) but hopefully that will change in the next couple weeks.  

I believe that while HashFast may not be persuing this their modules are very interesting from an immersion cooling perspective.  Flourinert is expensive up to $300 per gallon but a more useful way to look at it is the fluid cost per GH/s.  If you have 400 GH/s spread across 25 boards that is a lot of fluid and it may costs you >$2 per GH/s to immerse it and that isn't going to work long term.  If you have 400 GH/s on a single chip the fluid cost per GH/s can be much much lower.  I won't know how low for sure until I can get em into a tank but I estimate something on the order of $0.10 to $0.20 per GH/s in fluid costs based on the size of HashFast boards and the possible spacing.

... if you don't have the private key you don't have any Bitcoins (ask the users of Input.io, Bitcoinica, bitfloor, instawallet, etc, etc, etc).

I would say maybe a 3 PH/s seems likely (4 PH/s if KNC's Nov batch is towards the high end and others can pull together another Petahash. 
CT and HF are not shipping until Dec so for Nov it really just is BFL, Avalon, AsicMiner, Bitfury and all of those are pretty opaque but I can't see a massive surprise.  It just takes more and more haspower to move the difficulty (i.e. if BFL was shipping 0.2 PH/s per month and doubled that it would still "only" be 0.4 PH/s).  For Dec, HF first Batch is "small" looking like 0.5 PH/s (funny how half a PH/s is now small).  Cointerra is looking like 2 PH/s and HF second batch (and upgrades) is another 2 PH/s.  Dec really comes down to how much of that ships in Dec and how much slips to Jan.

You can start penciling in what you think 2013 is going to end.  Unless someone has a surprise waiting ....

Nice it looks like just the connectors are missing.  Finding a compatible connector on digikey shouldn't be too tough.  You don't by any chance have a caliper do you?   Getting the pitch (distance between holes) is all that is needed.

Anyone have a photo of a board with only 4 ports?  If it is just the physical connector is not present, they are through hole components so those handy with a soldering iron should be able to add the component.  If they are a different PCB which doesn't even have the traces and holes for the extra 2 ports or lots of other components are absent well that is a different story.

Agreed.  The Nov preorders will be two orders of magnitude more.  It is likely the "left overs".  They ordered so many hosts, cases, etc, used their inventory of parts to build the Sept/Oct/Nov orders and they came up with 200 extra boards.  Sell off the lot.  Why not? The value of ASICs goes down over time.  Why wait for the next batch just sell them off for less?

Your example is too crude to provide anything to correct.

First of all if you are mining for a pool you really aren't a "miner" you are a SHA-2 hashpower provider (independent contractor), your pool makes sure there is no duplicate work.  For the purpose of this explanation, only entities which construct the block headers are "miners" (pool operators, solo miners, and some unique situations like p2pool).

No miner is working on the same "work".  You aren't just hashing a single number.  That would be beyond pointless it wouldn't provide any security.  What does the number 123456 hashed mean?  Absolutely nothing.  How would it prevent modifying the blockchain?  It wouldn't.  

You are hashing a blockheader which consists of 6 values only one of which is "the number" which is called a nonce.  The nonce is simply a way to add entropy so you can make multiple tries without changing anything else in the blockheader.

So what is in the blockheader?

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

If ANY of these values are different the hash will be different.  Miners routinely attempt the same nonce.  Your own individual miners will attempt the same nonce millions of times.  The nonce is only 32 bit, there are only 4 billion possibilities.  A KNC Jupiter for example will exhaust the nonce range in less than 1/100th of a second.  Even a low end GPU will try all possible nonces (for a given blockheader) in a few seconds.

The "hashMerkleRoot" is a unique hash which represents all the transactions in the block.   If you add, remove, or change ONE transaction in the block you will end up with a different Merkle Root hash.  One of those transactions is the "coinbase" where miners receive their reward.  Even two miners (solo miners, pools, p2pool users) which have the exact same set of other transactions will still have a unique coinbase txs and thus have different Merkle Root hashes.

Different Merkle Root Hash = Different Block Header = Different Hash.

No two miners on the planet are ever working on the same work (with the exception of a bug in a pool server issuing duplicate work to its workers).

The host on some Jupiters has 6 ports but some only have 4.  Not sure if it is "early" vs "late" shipping or just random but if someone has only a rig with only a 4 port host it is not possible.  If you have a 6 port host there are reports in this thread of it working fine with 5 or 6 modules connected (one user took 2 boards from a Saturn to make a "Super Jupiter). Obviously you will be using more power so the power supply should be raised.  At least 1000W (on 12V rail) for 5 modules, 1250W for 6 modules.

IIRC the photos of the hosting "shelves" had standoffs for 6 modules.  So it is possible, if they will ever offer it is another question.

Jut curious but does bitcoind support IP blacklisting.  Yeah it probably can (and would be better done) at the firewall level but this topic just made me wonder if there is any config option in bitcoind.

Well if you resused the password then you can assume it was compromised somewhere.  Expensive lesson but never reuse passwords for anything financial related.   If you want to be lazy and reuse the same password for steam games and twitter that is one thing but anything financial related should be unique.

You indicated your system came back clean so:
1) Do you have a backup of your wallet.dat?  Is the backup stored online?  Have you emailed the backup to yourself?
2) Does anyone else have physical access to your computer?

You said someone withdrew from your "account".  What wallet were you using?  Was it an online wallet or a local (QT client) wallet?

Here you go, the paper that started it all.  Don't expect concrete protocol details it was written about a year before the first client was released however it is a good reference to answer the "why" instead of the "how".  I would recommend reading it straight through, think about it, and then read it again.  Bitcoin's "way" of doing things is rather complicated and often the first reaction is "hey why didn't they just do xzy?" or " thats stupid you could just do abc instead for cheaper".  with further reading and research the "genius" of the solution starts to appear.

http://bitcoin.org/bitcoin.pdf

More on the nuts and bolts:
https://en.bitcoin.it/wiki/Introduction#Bitcoin_Basics

I left the prior post incomplete when I said Bitcoin doesn't work on the concept of balances by not saying what concept it does you.  Bitcoin works on the concept of inputs and outputs.  Understanding Bitcoin without this high level concept is pretty much impossible and often leads to all kinds of incorrect assumptions.  The whole Bitcoin system works on inputs and outputs.  All tx consists of inputs and outputs.  All inputs are the output of a prior tx.  An output is "spent" by creating a tx which uses it as an inut.  Outputs can only be spent once.  All outputs in the blockchain are either spent or unspent.  Since only unspent outputs can be used in future transactions the working set of possible inputs at any time is the set of current unspent outputs (UXTO).  This allows any node at any time to validate the integrity of any new tx is a very quick manner.  Remember miners are't the whole security model.  Miners simply force a consensus between two possible (but both valid) views of the network.  A model where all security relies on miners gives miners too much power.  In Bitcoin, all nodes in real time validate every single transaction.  Because the input of all txs are simply a collection of prior unspent outputs this can be done very fast and without ambiguity. 

Regarding Satoshi "himself", I fall into the camp that thinks it is more likely that Satoshi is a psuedonym for a team of people.  Bitcoin includes so many different concepts (cryptography, distributed networks, database, game theory, economics) that I find it difficult to believe that is was developed fully by a single person.  Anything is possible though.  I imagine we probably will never know.  I do agree they or he chose a memorable name for the history books.

Blockchain will get larger.   There are checkpoints which prevent rollback of the chain beyond a certain point but Bitcoin doesn't work on the concept of balances so there is no way to create a "current balance" and delete everything else.  Even if there was it probably can't be done on Bitcoin in a decentralized manner.

However most users will not run full nodes.  They will run SPV clients which download just the portions of the blocks they need in realtime as they need them.   Some users will always have to run full nodes but the security model of Bitcoin doesn't require 100% of users to run full nodes 100% of the time.


Bitcoin doesn't work using the concept of "balances" there are a large number of technical reasons I would recommend starting with the bitcoin wiki.  When your wallet shows "2.5 BTC" what is really means is I see according to the current blockchain that there are unspent outputs totalling 2.5 BTC for which I have the private key to spend.

Timestamps are generally next to useless in a decentralized network.  They provide no security.  I can sign anything with any timestamp.  If you (say a new user) receives it weeks from now prove to me I signed it at the timestamp indicated.  Bitcoin doesn't rely on timestamps for its security model (except in a limited fashion in difficulty adjustment but even then safeguards are included to limit the effect of any malicious timestamp).

If you haven't already and want a technical understanding of Bitcoin I recommend reading Satoshi's paper first and then doing some reading on the Bitcoin wiki.  You will not learn it all in a day, week, or probably a month.  Most users will probably never need to.

One clarification.

Bob's PubKey isn't in TX A.  The Hash of Bob's PubKey is in Tx A.  PubKeyHash.  
This type of transaction is called "Pay to PubKeyHash".  Alice doesn't know Bob's PubKey and doesn't need to know.  She is paying to the hash of Bob's PubKey (which can be computed from the address Bob provided).

In the input of Tx B Bob signs the (simplified) tx with his private key and includes a copy of his PubKey.

When other nodes validate tx B they, hash Bob's PubKey and ensure it matches that of the PubKeyHash in Tx A.  They find tx A by the using the TxId and Index in the input of tx B.  The Tx Id gives the hash of the tx and the index indicates which (of many) outputs is being "spent" here.  Once a node validates the Input of Tx B is valid (corresponds to a real output which hasn't been spent) and has the correct PubKeyHash the node will verify the signature. 

Well if you are choosing point then you are just drawing a line where you want it to be.  It is a pretty line.  I like that it is blue.  The complete lack of antialiasing give is a kinda retro vibe. Not much value beyond that.

https://en.wikipedia.org/wiki/Linear_regression

 

Or you can ignore people with glowing orange buttons who make unsupported claims.

Not even sure how to respond to that?  It certain is a first.  

I am still me (but then again I guess a hacker would say that).

What is the r squared for your regression or did you simply draw a line on a chart?

The OP isn't saying Bitcoin will do "this" simply showing that there the line drawn has a high correlation to the known historical price points.  In other words the chart of actual price points can be simplified/reduced to the line drawn.   If the trend continues (and it may not) it will generally follow the line drawn.  In other words MATH ROCKS.