Basically, the wallet will send different work to each miner.

Sold 2/3 introductory units.

From the market right now, yes

Certainly, just point them to the same wallet computer

That's $5 a coin Current price is nearly $4 .

Yeah, I thought of going case-less as well, however I wasn't sure how shipping would work. I chose cases with good ventilation and a side panel that can be removed, so non-reference-design cards can ventilate from at least two directions very well. 1250W Seagates, talk about longevity.

What I think could be a nice midpoint between what you propose and what might be possible to easily implement, as I have proposed before (and may work on in the future) is a coin based on an algorithm that has a step order, or some post-processing steps. Some of the steps would be entropy-reducing (not one-to-one functions) while others would be weak mixing functions, that just act to further complicate the algorithm.

The step order of the hashing algorithm would be determined by the, say, last 8 characters of the most recent block (or the retarget block). Some steps would have huge lookup tables, which would be memory-intensive. The point is that some weeks (or retarget periods), GPU mining would be easier than CPU mining. Others, CPUs would have an advantage.

I have actually already made the algorithm, and did so with a few other people for a programming challenge. However, if I eventually make this alternate currency, I'll likely tweak the algorithm more.

A miner, like cgminer, would simply also be sent the step order from the pool/solo miner, allowing for an algorithm that changes and, in result, is ASIC-resistant, while still being something that could be predictable, with steps made in an intelligent manner so that a program that writes a new algorithm every retarget doesn't make an extremely unbalanced algorithm.

Sadly, I don't know enough C++ (or C, or C# ...) to make this a possibility. It's an interesting idea, though.   

Generally speaking, Litecoin is the main alternate currency at this point, and so it is the most commonly-traded coin into other alts.

If only Alternate Currency price could be modeled with a statistical model accurately, trading would be so easy.

Unfortunately, each different Alternate currency has a different adoption pattern, many go dormant for months and then come back, some have a huge initial burst and then are abandoned, and some are adopted quickly by exchanges while others take longer.

Generally speaking, logic would prevail with mining 1 BTC a day vs x amount of Alt per day so the value of the Alt would be BTC/Alt in a time period. However, this model simply loosely models reality. If this were true, dustcoin.com's percents should all be around 100% (aside from merge-mineable coins).

Adoption and usefulness sadly aren't two sides of the same coin, but they do have some correlation.

It may be a good idea to graph current prices, past prices, and try to visually predict future patterns, however some people get heavily burned while doing this. Personally, I like to diversify my holdings between a lot of alternate currencies and Bitcoin itself. It is never a bad idea to keep a few thousand of each coin, so if one becomes huge you aren't kicking yourself. With feathercoin, all my holdings are just from coins that I mined at launch, but if I had not been on the early mining boat I would be probably converting around 8% of my Litecoins into feathercoins.

Best of luck, trading is really confusing!

Same thing happened (and is kinda still happening) on my post. People seem to be able to build 2x7950 or 2x7970 systems for less than the cost of the GPUs themselves, I don't quite understand how they get those numbers.

I've priced it out before, but here's my estimates for my line (likely similar to Crypto's)::
Motherboard: $80-90
RAM: $30
PSU: $110
CPU: $40-50
Graphics Card: $740 (I do 2x7950 builds)
OS: $100
Case: $90 (doesn't apply to Crypto)

Not sure how people do such cheap builds, but I would certainly love to see their parts list, I could really slash prices on my miners

I think they may be considering just the cost of the GPUs, perhaps they have some spare Wattage and PCI-E 16x sized slots. Not sure.

If an altcoin brings something new to the table, its adoption is easily-explained. For example, if a coin was released that combatted extreme gains by early adopters, the coin would server a niche, or a particular purpose.

Adoption of the "scam-coins" as some people like to call them (coins that just have a few variables changed) in my mind is simply market diversity. Just because Hershey's makes chocolate doesn't mean we don't like Nestle. Same with coins, All alternate currencies today are viable for just the simple reason that they split up a market further, which can be looked at as both good and bad. There's nothing wrong with these coins, and they are based on tried-and-true code.

Gisfrancisco has made the first order, will be putting together a tutorial soon with how to create your own miner too

Might want to be a bit careful with free shipping, maybe limit to US only or something. Shipping across the ocean can be many hundreds of dollars

In order to calculate what the affect of your mining will have on your elec. bill, I'd suggest grabbing a Kill-awatt, it measures power draw at the wall. Plug your unit into it, plug the killer into the wall, and find a good stable average. This will be in WATTs. Your bill likely states how much you pay per kWh (kilowatt hour). Divide W by 1000 to get your total kW.


Example: You pay $0.13 per kWh, and your miner pulls 960W at the wall. Your miner runs 24/7 for a month. How much do you pay in electricity?
0.13*0.960*24*30 = 89.856

Therefore, your 960W miner (probably talking 3x7970 for that) at a cost of $0.13 per kWh, would cost around $89.85-$89.86 per month in electric. You would probably get 1800-2000 kH/s out of that machine, which with litecoin mining would get you $541.37. Subtract electricity, then divide into the price of your unit to get projected (but not accurate at all, things change all the time!) ROI. Yay math!

As far as looking into the future, I personally see Litecoin hitting $5 in a month (if not a week!!) and hitting at least $10 over the summer. However, difficulty, I feel, will scale with this.

Personally, I would build either 3x3 7950 or 4x2 7950 rigs, if you are doing out-of-the-box mining (no case, aka) I would opt for the 3x3 7950, better deal for hardware, get around 12-18% more bang for your buck if you have good enough cooling and such.

Sounds like you know how overclocking works then Just follow some of the guides around the Alternate Cryptocurrencies forums, there are some good guides for doing the clocks yourself there

I wouldn't go much more than 3 cards in a rig simply to keep it manageable, however if you don't mind using risers and such you could easily get 4 cards in one system. Being that you used to own your own computer store, I'm sure you're more than versed in PSUs and such, just get an efficient one

Sounds like your question was probably more intended towards "is it profitable" not the "how" side of it, but short answer is right now it's very profitable to mine with GPUs, and as time goes on, while coins will get harder to mine, simple supply and demand economics (think of Bitcoin like a commodity . . .) show that supply stays the same while demand increases (and arguably, supply decreases as each person can mine less...) so the $200 of coins you might earn this week could be worth $2000 later on.

Congratz on that overclock, that's insane! xD

Personally I would split it a bit. With 5k I would throw at least a thousand into buying the currency itself, however you mentioned wanting to invest it all in mining.

You could get a 3x7970 setup for around $1900, although if you decide to go litecoin mine you may want to look into 7950s instead. 3x7950 would cost you around $1700-ish and put out around 1.8MH/s if properly tuned. You could grab three of these for 1700x3 = 5100 and make some great mining proceeds. If you want to buy ASICs, I would go with two BFL singles (50GH/s).

If you are planning to enter into Litecoin mining, keep in mind ASICs will not do this. As well, it will take some good tuning to get your cards to run optimally. Some people on the forums (myself included) sell pre-built miners that come already optimized for you.

If you were to go this route, you could get 4 machines, each with 2 7950s (so, for $60 less, you are talking about missing one 7950 as opposed to making the units yourself), tweaked to reach close to their best potential. That would run you around $5000 as well, assuming you live in the US.

I'm probably not gonna call people, but I like giving personal touches to businesses, it makes them more fun to run. I'm doing the same, I sell the units a bit above part cost, with a subsidy from my testing/burn-in period. As well, if you know some kind of language like Java or the C-family, you may want to look into making your own setup interface, that's what I'm doing so that it is as easy as possible for a person to sit down and start mining.

You can always use multiple PSUs, too.

Hey. Some people are gonna try to steer you clear from building your own machine by saying that GPU mining will soon become outmodded by ASICs, however there are still alternate cryptos (Feathercoin, Litecoin . . . ) that you can mine with your GPU that the impeding ASICs will not mine. An investment into $2000 worth of BTC should be a good investment, however if you have that kind of budget there is nothing wrong with having some mining fun as well.

You could build a machine that would hit between 1.5 and 1.9 MH/s (litecoin mininG) and 1.5 to 1.9 GH/s for bitcoin mining for your mentioned prices.

I'd look into snagging a simple MSi or ASUS board with 2 or 3 PCI-e 16x (size, not speed) slots. Grab a cheap Sempron 145, and make sure your mobo has the AM3+ socket. Unless you want to use it for a lot of gaming, consider going low-end on the hard drive and grabbing a 250GB magnetic, or perhaps a 128GB SSD. It's not a bad idea to leave a 2 or 3-gpu system out-of-box. My company sells mining units in PC cases to go with the idea of modular design as well as allowing for easy shipping, and easy deployment. However, if you are doing the build yourself, you may want to consider running your stuff out-of-box if you don't need to have it on a desk.

As for RAM, 4 GB would be plenty for mining, 8GB would convert it to gaming. For cards, if you intend to do scrypt mining at some point you may want to look into MSI Twin Frozr 3 GD cards, namely the 7950s. 7970s are a bit of a headache to do stuff with unless you really know your way around litecoin/NVC/FC mining .

For the PSU, get a PSU that is Bronze or above, the higher metals have better power efficiency. Always leave yourself headroom on your PSU, and try to plan your PSU for future plans (If you want to start with two cards, and decide to upgrade to three, it's nice to not have to fork over a bunch of cash to get your self another PSU! A good number is 300W per card. This is a bit high sometimes, but if you are doing good overclocks you can hit it. My 3x7970 machine draws 860-940W at wall and is an 80+SILVER.

Go cheap on the Optical drive if you intend to get one (unless you want fancy blu-ray stuffs), and make sure you get an extra SATA cable or two, you never know when a HD or optical won't come with one.

Good luck!

It's cool to see other people starting this industry up as well You may want to ask customers to post pics once they receive their units, that's what I am intending to do.

Cheers!