I am looking seriously to build an affordable rig based on Avalon chips

[flyonwall]

These prices are higher than BFL's, but much better than GPU cards. These will be standalone units, very easy to use. Connects to your router either through WiFi or Ethernet cable. Each unit includes its own microcontroller for user interface, communications, and implementing the most popular mining pool protocols. The only downside is that these will have to be pre-ordered also. I figure I can start shipping the lowest-cost unit by October, with third-party guarantee of delivery. You can choose to cancel your order and be refunded through third party, if your order is not delivered as promised. (Third party is not exactly like an escrow, but they have very good reputation.)

Again, these will be using Avalon's ASIC chips. However, Avalon has NOTHING  to do with this proposed project. I am simply buying chips from them to use in these units.

[1714469271]

1714469271

[cimpex]

october would be much too late i think.

[rttnpig]]

By october difficulty will be ridiculously high and it will be imposible to meet ROI

[flyonwall]

What do you think the difficulty will be by October? What is the current worldwide hashing rate?

Using the bitcoinx.com profitability calculator, even if we double the difficulty, it is still possible for the 1.12GH/s unit to earn more than $100 a month, AT CURRENT BITCOIN PRICES.

[flyonwall]

I used the following parameters in the calculator at http://www.bitcoinx.com/profit/

Bitcoin difficulty - 18,974,296
Hashrate - set this to 1120 (Mega hashes per second)
Power consumption - 6.6 watts
Cost of Mining Hardware - 265

[computerparts]

I used the following parameters in the calculator at http://www.bitcoinx.com/profit/

Bitcoin difficulty - 18,974,296
Hashrate - set this to 1120 (Mega hashes per second)
Power consumption - 6.6 watts
Cost of Mining Hardware - 265

Add 190 million to that difficulty and you should have an accurate guesstimate for October.

[flyonwall]

Hmmm. How did you come up with that guesstimate? The graphs here are not logarithmic, and so we can see the absolute increase in total network hash rate: http://bitcoin.sipa.be/

We can only extrapolate from the graphs, but isn't your guesstimate about a decimal place off?

[flyonwall]

Also, remember that BFL has just started shipping and they are fulfilling their orders slowly, and Avalon is shipping in slow batches. BFL is only now shipping orders placed last June, each with a capacity of 4 to 5 GH/s.

It is much, much easier to deliver chips than to build complete, tested systems.

[Amph]

I used the following parameters in the calculator at http://www.bitcoinx.com/profit/

Bitcoin difficulty - 18,974,296
Hashrate - set this to 1120 (Mega hashes per second)
Power consumption - 6.6 watts
Cost of Mining Hardware - 265

Add 190 million to that difficulty and you should have an accurate guesstimate for October.

i don't think, in 6 months difficulty will rise that much....

[bcpokey]

Honestly pre-orders are not viable at this stage of the game. I don't know what difficulty will be in october, and I don't want to say that it WILL be anything, but it seriously looks to be going up and up. There is little incentive for me to give all the risk on something that looks very unpromising.

New ASIC ventures need to take risk on themselves IMHO. Or at least split the risk, such as the burnin allten projects, where people buy the chips themselves (taking risk) but offering the service to put chips into something to make them go that will be READY TO GO at time of payment.

That's my opinion, for whatever it is worth.

[Gomeler]

If you are just now ordering your chips, Avalon already has something like 150-200TH of chips sold. Just those alone are enough to 3x-4x the difficulty. This doesn't include ASICMINER(who has plans for ~65 TH) or BFL. I fully expect difficulty to crest 50 million by December. In fact I imagine it'll be closer to 75 million by then.

[Amph]

x10 by the end of the year, look more reasonable, imho

[flyonwall]

If you are just now ordering your chips, Avalon already has something like 150-200TH of chips sold.

Is this info published somewhere? Even if true, we will all receive our reference board and communication protocol specs from Avalon at the same time, this May.

Can we assume that those chip orders are mostly for private use? If so, larger capacity boards will have to be designed and built for them: orders of magnitude more difficult than a circuit board with 4 chips and a micro-controller. Simulations will have to be done, and simulating a board with hundreds of chips on it is much more difficult than building a prototype for a 4-chip system. Manufacturing is also much easier and simpler for the 4-chip system.

If thousands of 4-chip systems come online ahead of 1000-chip systems, guess whose profitability is most affected. The nice thing with these small systems is that your capital exposure is very small compared to big systems. You can expand as your Bitcoin generation allows. If you don't generate enough Bitcoins, then you can't buy your next unit. However, if you are profitable and generate enough Bitcoins, you can buy as many units as you think can make you even more profitable. This is the Bitcoin miner for the common man.

[flyonwall]

New ASIC ventures need to take risk on themselves IMHO. Or at least split the risk, such as the burnin allten projects, where people buy the chips themselves (taking risk) but offering the service to put chips into something to make them go that will be READY TO GO at time of payment.

It would be good if we can do that, indeed. Problem is, these chips are not your garden-variety chips. Minimum order is for 10,000 chips at BTC 780 per order.

[computerparts]

Hmmm. How did you come up with that guesstimate? The graphs here are not logarithmic, and so we can see the absolute increase in total network hash rate: http://bitcoin.sipa.be/

We can only extrapolate from the graphs, but isn't your guesstimate about a decimal place off?

It's really not that difficult to speculate. My estimate is very conservative. We're talking October here. By then Asicminer will be fully deployed. Avalon's batches should be fully delivered and the chips should be fully assembled into boards. BFL should be completely caught up. There is a graph somewhere of BFL orders. There In that graph are 50 minirigs and much more other units. Add that up to the people who did not participate in that graph and there will be more. Then there's Kncminer who remains to be seen as legit or not but so far it is looking like they are. You are simply delusional to be estimating difficulty at only 18 million in October.

[flyonwall]

Yes, you're right. By October 1.12 GH/s won't be enough to be profitable.

I need to build circuit boards for those who are doing a group-buy of Avalon chips instead. Let me see how much interest there is on that.

[Dabs]

Difficulty is going to hit 1 billion. This year.
*edit.* at least it's going to hit 100 million before October.

It really depends on when all the ASICs arrive, if they arrive, when they get turned on and mining.