Thank you for the feedback.  This is precisely why I wanted to put out there an approach and some numbers before moving forward with an offering.


Difficulty will change in 2 days and go up almost 10%.  Specifically, after the change in difficulty, 9.3Gh/s will yield 39.5 btc/week.  Given the close proximity of the retargeting, I wanted to use that rate rather than the higher current rate.


Each rig has 3 x 6990's.  It would be awesome if they could only use 1kw each, but they use quite a bit more.  Based on my readings from a Kill-a-watt power meter, each card is using 366 watts of DC power.  The CPU and chipset use about 35 watts.  Thus, each rig is using about 1130W DC.  With 80plus Silver 1200W PSU (85% efficient at high load), the power at the wall is about 1330W.  4 rigs x 1330W= 5.3kw.  5.3 x 24 x 7 x $0.052 = $46 per week.  At $4.60, that's 10 btc/week, leaving a net of 30 btc per week.

However, power is not the only cost.  Assuming that we leave out depreciation, the rigs still need infrastructure - Internet connection, facility charges, and ventilation/cooling.  An honest accounting of costs should include this.  The costs are not particularly high, however.  I can share these costs with the other 60-ish rigs in my personal mining farm.  I expect the costs for these four rigs to be in the range of 1.5 btc/week.

Thus at the imminent difficult we would mine 40 btc/week.  Subtracing out 10btc in power and 1.5 btc for infrastructure costs = 28.5 btc after expenses.

If 30% is reserved, that results in 8.5 btc reserved and 20 btc to be distributed among shareholders as a dividend.  If there are 4000 shares, that's a payout of 0.005 (not 0.004 as I previously estimated).  By comparison OgNasty's MergedMining company (a well run mining company) in the most recent week paid out 6.22 btc across 1500 shares = 0.00414 per share.  To my knowledge he has no significant reserve for expansion.  Compared to other mining firms, I will be able to offer dividends equal to or higher than others while also reinvesting a large percentage. 


First, on the topic of CEO ownership, the CEO holding significant portion of the shares is not that unusual.  Bitcoin Syndicate, for example, holds 45% among the owners.  In this case I opted for less than 50% such that the shareholders would have controlling interest.

Regarding valuation and mining power, below are values for other mining companies:

Mining Co.	Shares	Share Price	Value	Hashrate (Gh/s)	BTC/Gh	Mh/share	Mh/btc invested
Bitcoin Syndicate	12,000	0.250	3,000	6.0	500	0.5	2.0
BMMO	4,000	0.357	1,428	4.0	357	1.0	2.8
MergedMining	5,500	0.133	732	1.6	457	0.3	2.2
Tygrr	1,500	3.000	4,500	12.0	375	8.0	2.7

Taking a look at my first proposal for Fresca, and a revised proposal:

Mining Co.	Shares	Share Price	Value	Hashrate (Gh/s)	BTC/Gh	Mh/share	Mh/btc invested
Fresca	4,000	1.000	4,000	9.3	430	2.33	2.33
Fresca (revised)	3,333	1.000	3,333	9.3	358	2.79	2.79

From the above numbers you can see that 4,000 shares at 1 btc is comparable to other mining firms.

After revisiting the original proposal, I am suggesting a revision to enhance the competitiveness of the offering.  I propose to reduce the total shares to 3,333 instead of 4,000.  The split would be ~50/50 between me and shareholders with 1667 offered for purchase by shareholders and 1666 held by me.   (I have updated the prospectus in the first post to reflect this proposal)

This would have the benefit of increasing the dividends per share.  Using the estimates from above, the payout would increase from 0.005 per share to 0.006, which I believe is quite high, while keeping the price at 1 btc/share.

Keep in mind that the company will retain 30% of earnings, which will be used for expansion, adding at least another 1.5Gh/s over 12 months.  That and the prospect of lower power prices will allow for greater profit potential in the coming months.

Thank you again for your feedback.  I would enjoy hearing your further comments.

Good question.   The primary goal of the 30% retained earnings is to mitigate variance with payouts.  Each week, based on difficulty and hash rate, there is an expected level of earnings.  Let's say it's 40 btc this week.  If we earned 42 btc this week, then the excess is put into the reserve.  If we have bad luck or a network failure and only earn 30 btc, then we can dip into that reserve so that weekly earnings match the expected level.  Either way you get the expected return of 40 btc each week.

Because we will be mining with DGM variance shouldn't be as much of an issue, but luck is still a factor, and hopefully we won't have many hardware issues.  Thus the reserve should grow over time to a point where its balance is more than is needed to ensure stable payouts.  That is the point in which we would want to look at adding new hardware and increasing the hashrate.

I've put together a draft of the new mining operation.

You can find it in this new thread:
https://bitcointalk.org/index.php?topic=74044.0

Many of you may have contributed to or followed my previous thread asking for input on a new mining operation
https://bitcointalk.org/index.php?topic=66204.0

I've had a chance to do some reflection on a strategy for a new mining company while incorporating many of your excellent ideas.  Below is a DRAFT of the IPO.  I would like your input before I finalize it.

Fresca Mining (Planned GLBSE symbol: Fresca) will produce comparatively higher returns by leveraging low cost electricity ($0.052/kwh) and dense GPU rigs for high profitability bitcoin and namecoin mining.  The company will issue weekly dividends but retain some earnings to grow the company.  The company will be able to share infrastructure such as switched PDUs, Cisco network equipment, and data center space with my personal mining farm to save costs and maximize profitability.

Initial hashrate: 9.3 GH/s (I have them running now as the user "Fresca" on maxbtc.com - note that they are currently running at stock speeds until I tune then, hence the lower hash rate)

Hardware (4 rigs as follows):
      
       (seen here as I set them up)

• MSI 890FXA-GD70 motherboard
• Antec TPQ-1200 1200W 80plus Silver PSU
• AMD Sempron 145 CPU
• G.Skill 1GB DDR3 RAM
• Kingston 4GB USB drive
• 3 x MSI or XFX 6990 GPUs
• 208v power via Server Tech Sentry switched PDU
• Gigabit Ethernet connection on Cisco managed switch
• Coolermaster 230mm fan

Dividends

• Dividends will be issued weekly
• Dividends will be calculated as net earnings divided by number of shares held
• Net earnings are earnings after retaining 30% for growth and costs of depreciation, power, internet, and facility
• At present BTC price and difficulty this would result in approximately 0.004 0.006 btc/share

Reinvestment and growth

• 30% of earnings will be set aside to ensure stable dividend payouts and to fund growth
• Acquisition of new hardware will be funded through reinvestment of retained earnings as opposed to issuance of further shares, unless otherwise decided by shareholders through a motion
• Future plans include the purchase of more rigs based on the most profitable technology at the time of purchase and a move to a new data center than offers power under $0.03/kwh

Policies and Records

• My goal is to be transparent in all finances.  I will keep track of finances in a Google Docs spreadsheet available to everyone
• Changes in company direction, dividend calculation, acquisition of new hardware, dissolution, or other important changes will be put to shareholder vote.  Because I already have a large mining farm of my own, I am happy to be very flexible with the direction of this mining company.

Ownership

• 4,000 3,333 shares will be created
• 2,000 1,666 shares will be held by me (50% - 1 share)
• 2,001 1667 will be available for purchase (50% + 1 share)
• Anticipated purchase price will be 1 btc

UPDATE 1:
I have reduced the number of shares while maintaining the share price in order to further enhance the competitiveness of this offering.

Thanks!

Excellent idea.  If I had some spare btc on hand I'd buy it. 

Great work Nefario!

Would you say that it's stable enough to launch an IPO or should a wait a bit for the kinks to be worked out?

I see how that phrase could be read a couple of different ways.  My power company has a two tier pricing model - small and large commercial customers.  The cut off between the two tiers is 50kVA of usage during a month.  Thus, once I am meeting or exceeding that level of power usage, then then rate will come down.  I did just speak with the power company and they told me that they review usage on an annual basis, not on a monthly basis.  So it might have to convince them to consider changing the rate without waiting for the annual review if I'm using at least 50kVA.

A better scenario is for me to move my operation to a different county where the base power rate is already lower.  Given that I'm nearing my power capacity at my current facility, that might be a necessary move.  Meanwhile, I'll have to stick with the 5.2 cents/kwh, which is still quite good.

I have an Intel Pentium D 805 CPU that I'd like to sell. Would be the perfect cheap dual core CPU for a socket 775 board

http://www.newegg.com/Product/Product.aspx?Item=N82E16819116001

CPU only - no heatsink

Asking 2 btc shipped

I now have about 200 GPUs.  It certainly is taking a lot of my time.  I've been able to use some tools like BAMT to make things easier to monitor and manage. I would like for this to become a full time endeavor by the end of summer, but I'm not quite in a position to do that yet.

More IPO details coming tomorrow...

Did you also try fans facing down - the way it would be if it was in a computer case?  Curious to know how that performs

Hi MentalFuneral and welcome to the forums!

I've been running many GPUs 24/7 since last July.  So far none have died completely.  I have had fans go out on about 10% of my GPUs during that time.  I have also noticed that the cards don't overclock as well over time.  For example, a 5830 that ran nicely at 960Mhz for many months is now running reliably only at 940 or even 920 MHz.  With a voltage increase you might be able to keep the clock rate high, but I prefer not to increase voltage.

I'll definitely try my best for a solid ROI.  That being said, changes in price and difficulty, mining luck, etc. will definitely keep me on my toes!

The data center is in Washington state - the cool Pacific Northwest where hydro power is plentiful and in the right places can be quite cheap.

For renting out processing power, I'm thinking first of password cracking, but honestly I've got my hands full with my bitcoin mining in the short term.  I'm planning to set up a test rig soon with backtrack and figure out an optimal configuration.  Rainbow tables are an option too, but haven't really spent much time fiddling with that yet.

Tom's hardware has its review posted.

http://www.tomshardware.com/reviews/geforce-gtx-680-review-benchmark,3161-2.html

Thanks for the link.  It came from the factory with a form of flexible fabric electrical tape.  I'll keep an eye on it and see how it runs.

Yup, I tried it and it works.  Both fans spin up at idle.

I have seen what you are talking about in terms of only one fan on at idle - not on this car - but another card.  Not sure why one would do it and not another.  But I guess as long as they are running when mining then it's all good.

Ok, I got it all put together and it works!     Here's the wiring:



And here it is all assembled - I like the Asus fans on the Gigabyte GPU.  haha

Haha.  Without the underlining of a link it looked like it was one link.  I would have never known if you hadn't mentioned it


That's what I figured.  I'll give it a try tonight and see what happens!

Cool.  I'll PM you

Thanks for the interest.  I've been a bit busy bringing online more equipment and working out what is the optimal starting point in terms of GH/s, strategy, and dividends vs. growth. Also, I've been waiting for GLBSE 2.0 to launch.  I hope to have everything worked out this weekend.