ASICMINER: Entering the Future of ASIC Mining by Inventing It

[minerpumpkin]

Good work! Although slightly too pessimistic, in fact!  
The June batch is supposed to be about 67 PH/s, thus costs about $13.4m. We can, in fact, pay for the batch with the revenue from the May batch. But you are right that - given a profit of $0.15/GH/s - we only make a surplus of $10.05m (revenue of $23.45m - production cost of $13.4m). If we paid that out in total, we'd get 0.0386 BTC/share. If we reinvested half of the profits, we'd be at 0.0193 BTC/share.
If we reinvested half of the total revenue, we'd be at 0.0225 BTC/share. The key is to find the sweet spot.

And July batch is 100+PH/s right? This sounds like the house bubble...

To be fair, it's the 'best' thing one can do. Try and reinvest as long as it makes sense. After it's not profitable anymore, stop and switch to gen 4 and/or pay out revenues.

[1713425405]

1713425405

[1713425405]

1713425405

[1713425405]

1713425405

[1713425405]

1713425405

[hdbuck]

So about 0.01 btc per share is probably about the most total dividend one could possibly expect for now, until gen 4.

you forgot about selfmining and franchisees.

If we assume that selling chips is the most profitable and fast(!) way to go, every KH/s that goes into self-mining and franchising, is a KH/s sold for less. i.e. we'd be off worse.

yea so even if those numbers you pulled out of your tinfoiled hat were close to reality, chip sales would only constitute first round of divs..

Did I really appear to be 'trolling' or 'attacking' AM over the past 6 months? Those are the soundest numbers we can get right now. I invite you to:

- Provide numbers that are more reliable or founded.
- Extrapolate from my calculation, how much dividends are going to accumulate over the next few months.

i understood its useless to come up with random numbers by now. only FC and board members know whats best for the sake of AM. i invite you to believe everything has been done for the best but without revealing all of it out to the competition or just sell your shares. (to me )

[minerpumpkin]

i understood its useless to come up with random numbers by now. only FC and board members know whats best for the sake of AM. i invite you to believe everything has been done for the best but without revealing all of it out to the competition or just sell your shares. (to me )

You are right, this is usually a truly futile endeavour! But now we have some concrete numbers we can use to try and valuate the company. And it's dangerous for everyone invested to ignore those numbers and say "we can't be sure".

Imagine a trail of cookie crumbs leading to a child's room. The child's brother also testifies "Yeah, I heard him/her going in the kitchen, but I don't know whether he/she took a cookie". We can't know for sure whether the child actually took a cookie. Yet, all the clues we have point in that direction. Continuing to be completely certain - or hoping - that there is still the same number of cookies in the cookie jar would be naïve.

[minerpumpkin]

Disclaimer: I don't own any of these shares anymore, but am tempted to pick some up again.

My calculation/speculation:

- Cashflow is near zero as of May 27
- Liabilities are near zero
- So NRE-costs for gen 3 are paid for
- May batch of 33.5PH is paid for: Point "Materials" in the balance sheit is ~6.1 Million USD  ( ~33,500,000 GH*~0.2 USD/GH ). Really curious though if "Materials" are really chips and what "Products" translates to

1) What is the status, size, and expected delivery of the next batch of chips? What about the one after that?
re 1) This month: 850k, next month: 3.35m (order size), June: 6.7m (order size), assuming each chip is 10G.

- May batch is being sold for ~0.35USD/GH, though I don't really believe the following source...0.35USD/GH just seems way too low. If they really sell that low, most 28nm chip-producers will seriously have tough times coming up

http://www.cybtc.com/forum.php?mod=viewthread&tid=7951&highlight=%E7%83%A4%E7%8C%AB

- So revenue for May batch: 33,500,000 GH * 0.35 USD/GH = 11,725,000 USD. One third of this may be retained for the next batch, leaves ~7.8m USD for the May batch-dividend. Translates to ~0.0296 BTC per share @ 660USD per BTC.

(BoardMember1) so I underatand that the next payments to the shareholders shoudl happen in May....
(BoardMember2) regarding dividends, do you have a target for retained earnings vs. dividends?
(BoardMember2) if retained earnings is high we need to consider the security mechanisms for the company treasury.
(FC) At most 1:2. Most retained are for short-time spendings (license, wafer/components order). There are no long-time plan of retaining since we won't buy land/bond/etc.
(BoardMember2) Alternatively AM could get "invested" in related businesses
(BoardMember2) If no opportunities exist I agree to an aggressive dividend schedule

The dividend schedule will be aggressive, as AM will not require large sums of retained capital. The rationale is that AM doesn't need to invest large sums into infrastructure and that the majority of chip production costs will be covered by the business partners on a contract schedule.

- If the June batch (67PH) can really be paid for by just 1/3 of the May batch-revenue plus the contract schedule with business partners and will sell for 0.3 USD/GH, we're looking at 33.5 USD per share ( I don't dare to speculate on the USD/BTC-ratio in July )

Please correct me if I'm wrong somewhere, thanks in advance. Also @ the spoondoolies-fans: Do you know for how many dollars 1GH of SP30 chips is going to be produced?

I didn't double-check your calculations but it seems to be a sound analysis! Well done!
- I even do think it may be possible for the May batch to be sold for $0.4. Paying for the June batch with a third of the revenue still seems to be possible in your calculation (we may be short $1m or $2m, but effectively this works).
- You are right that we don't know about the BTC price in July, but it probably won't go down right now and thus will be cutting dividends. I doubt that we can sell chips at a significantly higher price then, though - given a steady BTC price increase and ramping up of competitors.
- Everyone may now try and extrapolate from there how many months we are able to continue this cycle while the price we are able to sell the chips approaches the fabrication price. (remember AM is said to have a 30PH/s/month capacity in creating wafers https://bitcointalk.org/index.php?topic=99497.msg6914171#msg6914171)

A neat calculation!

[Another User]

I used to be very bullish as well, but do acknowledge that the expectations and numbers from January (0.35J/GH/s, March delivery, weaker competition) need to be re-evaluated. I invite you to present a calculation based on those new premises.

An excellent idea because what happening here is that you are lowering the numbers of profit margins (which is fair) but keeping the costs at a fixed amount (which is not always true). Production costs for wafers/chips, at a general understanding, decreases overtime on older, larger die sizes AND varies on the quantity produced. As quoted from a previous friedcat post, the going rate was "<$0.2/GH" at the time (Nov '13) and aimed for March '14 delivery. If the "costs" go down, the savings can be passed on the buyers via a lower rate and providing incentives for more purchases.

It has been stated that chip sales will continue until it was more profitable to mine and resources would be adjusted accordingly. Now, if the margins have dropped due to various mentioned reasons, wouldn't the mining start already and the sales stopped?

Because it's too much fun, I'll humor your scenario: Let's say that the production costs are fixed at the $0.2 a GH, as you say, and they decide to continue flooding the market with decreasing profit margins in the coming months. That then eliminates any potential franchisees with such a high network rate (wasted funds on immersion cooling )and any revenue from self-mining (no moar GH/s for me?  ). At this going rate, chip production can't continue and AM is bankrupt before Gen4 can even begin (about Aug/Sept '14 in this scenario). GG Bitfury & Spondoolies. But hey, you're right: it's not a pretty picture when you're using estimated Nov '13 costs with estimated April/May '14 profit margins. I would dump my shares too if friedcat was that short-sighted.

Now the question we are all wondering now is what has been the production costs of April & May and will be in June & July. We're not going to get an answer to that but it is safe to extrapolate here that if they can afford to have lower prices, it is due to the lower production cost. Otherwise friedcat = big dummy.

Don't get me wrong, I get what you're doing and I'm picking up the cheap shares as well 

[minerpumpkin]

I used to be very bullish as well, but do acknowledge that the expectations and numbers from January (0.35J/GH/s, March delivery, weaker competition) need to be re-evaluated. I invite you to present a calculation based on those new premises.

An excellent idea because what happening here is that you are lowering the numbers of profit margins (which is fair) but keeping the costs at a fixed amount (which is not always true). Production costs for wafers/chips, at a general understanding, decreases overtime on older, larger die sizes AND varies on the quantity produced. As quoted from a previous friedcat post, the going rate was "<$0.2/GH" at the time (Nov '13) and aimed for March '14 delivery. If the "costs" go down, the savings can be passed on the buyers via a lower rate and providing incentives for more purchases.

True, the costs of producing the chips are declining as well, sure! But: The difficulty and competition are ramping up on a much higher pace. You can run from a train, sure, but you'll only gain a few meters. We don't know whether the price of $0.2 was indeed for November '13 or the planned production in early '14. Anyways, 40nm technology isn't cutting edge anymore and I highly doubt the cost per GH/s will retract below $0.15 in the following months or even next year. It'll prolong the viability a bit, sure - but I assume this really is negligible.

It has been stated that chip sales will continue until it was more profitable to mine and resources would be adjusted accordingly. Now, if the margins have dropped due to various mentioned reasons, wouldn't the mining start already and the sales stopped?

No, of course selling hasn't stopped. It is still the best way to go. I never said gen 3 wasn't viable, it just seems not as viable as we all (including myself) hoped for.

Because it's too much fun, I'll humor your scenario: Let's say that the production costs are fixed at the $0.2 a GH, as you say, and they decide to continue flooding the market with decreasing profit margins in the coming months. That then eliminates any potential franchisees with such a high network rate (wasted funds on immersion cooling )and any revenue from self-mining (no moar GH/s for me?  ). At this going rate, chip production can't continue and AM is bankrupt before Gen4 can even begin (about Aug/Sept '14 in this scenario). GG Bitfury & Spondoolies. But hey, you're right: it's not a pretty picture when you're using estimated Nov '13 costs with estimated April/May '14 profit margins.

Self-mining, franchised mining etc. makes only sense when we can't sell chips anymore. The turnaround time for selling chips is much faster. Every KH/s used for mining instead of selling chips (as long as it's profitable) is a KH/s wasted. I do agree though, that mining is good thing, as it festers independence and a continuing stream of income. But effectively it means less income/dividends.

I would dump my shares too if friedcat was that short-sighted.

Well I guess he just ran into too many problems (delays, surprise power consumption, competitors) and additionally the expectations from gen 1 (6x returns, 4 BTC share price) were way too high. FC didn't do it on purpose (Hanlon's Razor) but I also assume that it was just bad luck, rather than stupidity or negligence. I do think FC is pretty smart!

Now the question we are all wondering now is what has been the production costs of April & May and will be in June & July. We're not going to get an answer to that but it is safe to extrapolate here that if they can afford to have lower prices, it is due to the lower production cost. Otherwise friedcat = big dummy.

Indeed! A very very valid question! But I don't think that they're below $0.15/GH/s anytime soon, and thus only negligibly smaller. Feel free to re-run the numbers!

Don't get me wrong, I get what you're doing and I'm picking up the cheap shares as well 

Actually, I've decreased my position dramatically for the time being. AM is still viable, but won't be able to issue dividends as high as people around here still expect (I was very disappointed as well, believe me!).

[whalezy]

1) What is the status, size, and expected delivery of the next batch of chips? What about the one after that?
re 1) This month: 850k, next month: 3.35m (order size), June: 6.7m (order size), assuming each chip is 10G.

Still don't understand the financial report, in the cash flow sheet, the chip sales income is around $4.5m, divided by $0.35/G, means the sales in May was 12.85PH, it's quite below 33.5PH. AM I wrong about this?

[vortex1878]

1) What is the status, size, and expected delivery of the next batch of chips? What about the one after that?
re 1) This month: 850k, next month: 3.35m (order size), June: 6.7m (order size), assuming each chip is 10G.

Still don't understand the financial report, in the cash flow sheet, the chip sales income is around $4.5m, divided by $0.35/G, means the sales in May was 12.85PH, it's quite below 33.5PH. AM I wrong about this?

Any attempt of interpretation of the sales figures in the cashflow-statement is just wild speculation as we do not know the payment terms with the customers.

[minerpumpkin]

1) What is the status, size, and expected delivery of the next batch of chips? What about the one after that?
re 1) This month: 850k, next month: 3.35m (order size), June: 6.7m (order size), assuming each chip is 10G.

Still don't understand the financial report, in the cash flow sheet, the chip sales income is around $4.5m, divided by $0.35/G, means the sales in May was 12.85PH, it's quite below 33.5PH. AM I wrong about this?

The chip sales income is most likely income from the April batch, maybe some May - we don't know. This gives us about 10 PH/s for April. The majority of the May batch is supposed to be in the inventory, which is probably priced in production costs.

[buck19]

Self-mining, franchised mining etc. makes only sense when we can't sell chips anymore. The turnaround time for selling chips is much faster. Every KH/s used for mining instead of selling chips (as long as it's profitable) is a KH/s wasted. I do agree though, that mining is good thing, as it festers independence and a continuing stream of income. But effectively it means less income/dividends.

I agree that selling chips is a way to get the maximum profit up front versus mining that will spread the income out over time. However, in this climate of an escalating USD/BTC exchange rate, mining with the chips eliminates the exchange risk.

If USD/BTC rises to $5000, our bitcoin profit from chip sales will become miniscule, whereas a mining farm will still be happily cranking out the coins.

A balance should be struck between sales and self-mining.

[whalezy]

1) What is the status, size, and expected delivery of the next batch of chips? What about the one after that?
re 1) This month: 850k, next month: 3.35m (order size), June: 6.7m (order size), assuming each chip is 10G.

Still don't understand the financial report, in the cash flow sheet, the chip sales income is around $4.5m, divided by $0.35/G, means the sales in May was 12.85PH, it's quite below 33.5PH. AM I wrong about this?

The chip sales income is most likely income from the April batch, maybe some May - we don't know. This gives us about 10 PH/s for April. The majority of the May batch is supposed to be in the inventory, which is probably priced in production costs.

Thanks, so the materials refer to wafers, assuming its cost is $0.15/G. Products refer to chips assuming its cost is $0.2/G. Therefore, we have around 48PH in hand, and will generate revenues as 16.8m($0.35/G). There are still 50PH is waiting to be produced. Minus the cost 10m, the remaining profit is around 6.8m, in current btc rate, per share dividends for this batch is 0.0257btc. The last 50PH will generate 15m($0.3/G), let's assume AM will retain half of them and cease Gen3 production, the dividends for the last batch is 0.0284btc. TOTAL 0.0541. Unfortunately, we are in a btc bull market, that may be the biggest risk.

[laustcozz]

I don't understand why no one gets that the price of BTC going up is a good thing.  The higher the price of BTC the more we can charge per H/s.  The price we can charge will increase proportionally while our costs remain fixed, thus more profit.  BTC price increased by a bigger percentage last month than difficulty did.  That means we should be able to sell chips for MORE now than we could a month ago (in dollars).

[vortex1878]

Of course the sales price can always be adjusted for future sales.

Regarding that rumoured near-future 0.35$/G:

Anybody knows at what G-rate of the chip that price might be based on?

If it is the initial 12.8G/chip (and as now everybody incl. fc is talking about 10G/chip), the actual price would be 0.448$/G (0.35$ for actual 0.78G at 10G/chip).

Just a thought...

[minerpumpkin]

I don't understand why no one gets that the price of BTC going up is a good thing.  The higher the price of BTC the more we can charge per H/s.  The price we can charge will increase proportionally while our costs remain fixed, thus more profit.  BTC price increased by a bigger percentage last month than difficulty did.  That means we should be able to sell chips for MORE now than we could a month ago (in dollars).

That's been addressed as well. Like I already said, we're entering a more saturated market now - in comparison to last year. The competition is a lot stronger now and has good chips. It'll be a lot harder to re-adjust the price because the competition has to do so, as well. If we try to sell more chips, we may soon be limited by the maximum number of chips we may produce per month. Also, price adjustments often can't reflect the increase in BTC price that fast and customers may not be willing to buy a proportionally more expensive product.
But the most important thing is: If we sell chips today and encounter another bubble or dramatic increase in BTC price, we won't have sold more, only our divs (priced in BTC) will get slashed due to the exchange rate.
This is indeed one of the few cases, where mining makes sense and also offers increased stability.

[minerpumpkin]

Of course the sales price can always be adjusted for future sales.

Regarding that rumoured near-future 0.35$/G:

Anybody knows at what G-rate of the chip that price might be based on?

If it is the initial 12.8G/chip (and as now everybody incl. fc is talking about 10G/chip), the actual price would be 0.448$/G (0.35$ for actual 0.78G at 10G/chip).

Just a thought...

I'm not following... We're talking about $0.35 per GH/s, that is per GH/s. That Chinese offer seems to be our very own gen 3 chips, since it says "grilled cat 3 generations chips". Delivery seems to be June 10-20.

[vortex1878]

Of course the sales price can always be adjusted for future sales.

Regarding that rumoured near-future 0.35$/G:

Anybody knows at what G-rate of the chip that price might be based on?

If it is the initial 12.8G/chip (and as now everybody incl. fc is talking about 10G/chip), the actual price would be 0.448$/G (0.35$ for actual 0.78G at 10G/chip).

Just a thought...

I'm not following... We're talking about $0.35 per GH/s, that is per GH/s. That Chinese offer seems to be our very own gen 3 chips, since it says "grilled cat 3 generations chips". Delivery seems to be June 10-20.

Please think again. Or is my English really that bad?

EDIT:
I'll try again:
Yes, yes, of course we are talking about AM Gen3 chips here.
If the advertised 0.35$/G are based on 12.8G/chip, then at only - I assume more efficient - 10G/chip (and that is the latest official rate from fc) the actual price would be 0.448$/G.

[hdbuck]

Like I already said, we're entering a more saturated market now - in comparison to last year. The competition is a lot stronger now and has good chips.

saturated? besides spondoolies that may be doing ok (yet there is no comparison in term of quantity, nor in business model), i dont see a single ASIC manufacturer that offer stable, affordable and instock products. bitmain's S2 are a total disaster, avalon and hashfast i wont even talk about them, bitfury seems to be on the go now with that recent funding (yet its only the FUNDING phase and nothing really came out anyway) etc etc.. for the last year, the ASIC market has been ruined and undermined by scammy commercial practices and chip design amateurism with all of them wanting to produce 28nm chips straight on & at all cost. AM is here to change that.

sure current btc bull market is quite worrying, but FC mentioned OTC channels in HK. plus if bitcoin price rise, chips price to mine them will follow too. never in hell they'll sell at lost. fuck even BFL managed to put their 5Gh jalapenos in wallmarts for 300$!

[minerpumpkin]

Of course the sales price can always be adjusted for future sales.

Regarding that rumoured near-future 0.35$/G:

Anybody knows at what G-rate of the chip that price might be based on?

If it is the initial 12.8G/chip (and as now everybody incl. fc is talking about 10G/chip), the actual price would be 0.448$/G (0.35$ for actual 0.78G at 10G/chip).

Just a thought...

I'm not following... We're talking about $0.35 per GH/s, that is per GH/s. That Chinese offer seems to be our very own gen 3 chips, since it says "grilled cat 3 generations chips". Delivery seems to be June 10-20.

Please think again. Or is my English really that bad?

I thought you weren't sure which chips were being sold here: http://www.cybtc.com/forum.php?mod=viewthread&tid=7951&highlight=%E7%83%A4%E7%8C%AB.
Still not following. We're pricing the chips per GH/s. If 1 GH/s costs $0.35, a 10 GH/s chip costs $3.5 and the initial 12.8GH/s chip would be $4.48. But not $0.448/GH/s!?
The chips could do 50GH/s, they'd still be $0.35/GH/s.

10 GH/s may be due to the increase in consumption or heat they encountered, so the recommended speed has been cut to about 10 GH/s. The only thing that does happen is that the production price would increase by the factor 1.28x per GH/s, since one chip yields less GH/s.

[minerpumpkin]

Like I already said, we're entering a more saturated market now - in comparison to last year. The competition is a lot stronger now and has good chips.

saturated? besides spondoolies that may be doing ok (yet there is no comparison in term of quantity, nor in business model), i dont see a single ASIC manufacturer that offer stable, affordable and instock products. bitmain's S2 are a total disaster, avalon i dont even talk about em, bitfury seems to be on the go now too but nothing really came out anyway etc etc.. for the last year, the ASIC market has been ruined and undermined by scammy commercial practices and chip design amateurism with all of them wanting to produce 28nm chips straight on & at all cost. AM is here to change that.

sure current btc bull market is quite worrying, but FC mentioned OTC channels in HK. plus if bitcoin price rise, chips price to mine them will follow too.

Bitfury just got a massive cash infusion and they're quite talented in doing effective designs in low-nm architectures. I really do think we shouldn't underestimate Spondoolies. They seem to know what they're doing and their next generation could be quite powerful.

OTC channels are a great solution! They won't offer a different price, though. I'm just worried that the price of the chips won't be adjusted as fast as the BTC price, additionally money needs time to trickle in and be converted. A month is an eternity in Bitcoin bull/bubble-world.

[vortex1878]

Of course the sales price can always be adjusted for future sales.

Regarding that rumoured near-future 0.35$/G:

Anybody knows at what G-rate of the chip that price might be based on?

If it is the initial 12.8G/chip (and as now everybody incl. fc is talking about 10G/chip), the actual price would be 0.448$/G (0.35$ for actual 0.78G at 10G/chip).

Just a thought...

I'm not following... We're talking about $0.35 per GH/s, that is per GH/s. That Chinese offer seems to be our very own gen 3 chips, since it says "grilled cat 3 generations chips". Delivery seems to be June 10-20.

Please think again. Or is my English really that bad?

I thought you weren't sure which chips were being sold here: http://www.cybtc.com/forum.php?mod=viewthread&tid=7951&highlight=%E7%83%A4%E7%8C%AB.
Still not following. We're pricing the chips per GH/s. If 1 GH/s costs $0.35, a 10 GH/s chip costs $3.5 and the initial 12.8GH/s chip would be $4.48. But not $0.448/GH/s!?
The chips could do 50GH/s, they'd still be $0.35/GH/s.

10 GH/s may be due to the increase in consumption or heat they encountered, so the recommended speed has been cut to about 10 GH/s. The only thing that does happen is that the production price would increase by the factor 1.28x per GH/s, since one chip yields less GH/s.

Now I know it is not my bad English. ^^

I'll give it a last try with a bold example:
Somebody is offering you chips for 1$/G. He says each chip performs 1Gh/s (although the official rating is different).
At 1 Gh/s the chip consumes 10 Watts. But the most efficient (and rated) rate for the chip is 0.5Gh/s as it is only consuming 2 Watts then. So you might want to run the chips at the rated 0.5Gh/s only. Hence, you actually pay 2$/G.