Author

Topic: ASICMINER: Entering the Future of ASIC Mining by Inventing It - page 383. (Read 3917543 times)

hero member
Activity: 686
Merit: 500
A pumpkin mines 27 hours a night
Ahh, well there's a flaw in your reasoning, though: Producing the chips will cost effectively the same in 3 months or even 6 or 12. A fabrication process in that size isn't prone to such drastic price decreases (compared to difficulty increase) AFAIK. So they'd best be producing as many chips as they can sell or put to use within a reasonable timeframe without parking to much liquidity in chip orders.

OK. but the production costs of pretty much all technologies reduces over time as improvements in manufacturing technologies and competition increases. This is pretty standard in nearly all tech related industries.

I understand what your saying, but there should be at least a 5% drop in costs minimum over three months, and given production of this size, I would have thought it would be beneficial to fabricate allot closer to delivery times.

But at the same time, my experience comes from manufacturing technologies in general, this is my first foray into chip manufacturing. Are you certain? It just doesn't mix well with what I know about the manufacturing industry, especially in the digital space.

Thanks for your critique.

I'm not an expert in IC design or similar as well, but even if we saw a 50% decrease over 3 months, we'd still see a 500% increase in mining difficulty. Even if the production cost goes down significantly, it still can't 'outrun' the difficulty increase at all. I believe FC knows best and plans accordingly Smiley
sr. member
Activity: 800
Merit: 250
sr. member
Activity: 253
Merit: 250
CHRIST I'M EXCITED. Should buy my 100 shares now or wait for it to dip a lil more?
legendary
Activity: 817
Merit: 1000


I have waited for this day since I bought my first stupid share at 3 btc on the way down... I scooped up more then my fair share at 0.3 to make up for it:D
legendary
Activity: 1512
Merit: 1012
Still wild and free
legendary
Activity: 1260
Merit: 1002
...
I remember that discussion. To quote myself back in the days...
...
I'm AM bullish, but expecting (a lot) more than 0.5 BTC divs/share for all of gen 3 is delusional these days. Looking at the competition, this is still more than fine. Prepare for 0.25 BTC divs/share and take any surplus as a nice bonus. Disclaimer: This is no trading advice, remember that I may be very wrong about this. Also, I do own shares.

yea alright.. ^^
anyho, 1 week left! Cheesy
*drumroll
copper member
Activity: 224
Merit: 100
...as I'm assuming GEN 4 will be available in three/four months....

No way! First of all, gen 3 is viable and will have a longer lifespan than that. But more importantly, friedcat stated that anything below the process of gen 3 will require a lot more work, simulation, etc. He also stated in his April QA that gen 4 will be far more challenging. Even if they're working on it right now, the question will be whether there'll be a tapeout this year. I'd bet on no.

How about another round of questions after the first upcoming dividend? Minerpumpkin, would you volunteer to collect and forward the questions again?

Yes please. After doing my due diligence I have allot of questions.

Would be great if they just hired someone to do PR. they have more than enough revenue
full member
Activity: 141
Merit: 100
...as I'm assuming GEN 4 will be available in three/four months....

No way! First of all, gen 3 is viable and will have a longer lifespan than that. But more importantly, friedcat stated that anything below the process of gen 3 will require a lot more work, simulation, etc. He also stated in his April QA that gen 4 will be far more challenging. Even if they're working on it right now, the question will be whether there'll be a tapeout this year. I'd bet on no.

Yes, it seems that gen4 will take a lot more work than gen3, but we also have no idea when work started on gen4: Maybe AM was working on gen3 and gen4 simultaneously.
It would be good to have a rough estimate for the tape-out date of gen4, since right now we have absolutely no idea. It could be in two months, or in more than a year.

How about another round of questions after the first upcoming dividend? Minerpumpkin, would you volunteer to collect and forward the questions again?
copper member
Activity: 224
Merit: 100
The key aspect when they can't sell the chips is: self mining / franchising. They get the chips for free practically nothing. Selling the chips is the first priority, as it will yield as much as possible in as little time as possible. Keeping them in stock isn't very wise. So they'll be best off putting them to work for themselves.
Selling half of the chips seems like a reasonably conservative enough guess.

Agreed. To further your point, I don't believe they would fabricate that many now without knowing they could sell them quickly, as why fabricate that many at a price set now, when they could fabricate them in three months for a much cheaper price.

Ahh, well there's a flaw in your reasoning, though: Producing the chips will cost effectively the same in 3 months or even 6 or 12. A fabrication process in that size isn't prone to such drastic price decreases (compared to difficulty increase) AFAIK. So they'd best be producing as many chips as they can sell or put to use within a reasonable timeframe without parking to much liquidity in chip orders.

OK. but the production costs of pretty much all technologies reduces over time as improvements in manufacturing technologies and competition increases. This is pretty standard in nearly all tech related industries.

I understand what your saying, but there should be at least a 5% drop in costs minimum over three months, and given production of this size, I would have thought it would be beneficial to fabricate allot closer to delivery times.

But at the same time, my experience comes from manufacturing technologies in general, this is my first foray into chip manufacturing. Are you certain? It just doesn't mix well with what I know about the manufacturing industry, especially in the digital space.

Thanks for your critique.
hero member
Activity: 686
Merit: 500
A pumpkin mines 27 hours a night
The key aspect when they can't sell the chips is: self mining / franchising. They get the chips for free practically nothing. Selling the chips is the first priority, as it will yield as much as possible in as little time as possible. Keeping them in stock isn't very wise. So they'll be best off putting them to work for themselves.
Selling half of the chips seems like a reasonably conservative enough guess.

Agreed. To further your point, I don't believe they would fabricate that many now without knowing they could sell them quickly, as why fabricate that many at a price set now, when they could fabricate them in three months for a much cheaper price.

Ahh, well there's a flaw in your reasoning, though: Producing the chips will cost effectively the same in 3 months or even 6 or 12. A fabrication process in that size isn't prone to such drastic price decreases (compared to difficulty increase) AFAIK. So they'd best be producing as many chips as they can sell or put to use within a reasonable timeframe without parking to much liquidity in chip orders.
copper member
Activity: 224
Merit: 100
...as I'm assuming GEN 4 will be available in three/four months....

No way! First of all, gen 3 is viable and will have a longer lifespan than that. But more importantly, friedcat stated that anything below the process of gen 3 will require a lot more work, simulation, etc. He also stated in his April QA that gen 4 will be far more challenging. Even if they're working on it right now, the question will be whether there'll be a tapeout this year. I'd bet on no.

Ahhh ok. This still doesn't change much. I accounted for only selling fifty percent of the chips produced in the next three months, I guess we will just order more if we have too after that and sell them for a cheaper price. I do doubt though that they would fabricate that many chips without an idea for selling at least half within three months of delivery

The key aspect when they can't sell the chips is: self mining / franchising. They get the chips for free practically nothing. Selling the chips is the first priority, as it will yield as much as possible in as little time as possible. Keeping them in stock isn't very wise. So they'll be best off putting them to work for themselves.
Selling half of the chips seems like a reasonably conservative enough guess.

Agreed. To further your point, I don't believe they would fabricate that many now without knowing they could sell them quickly, as why fabricate that many at a price set now, when they could fabricate them in three months for a much cheaper price.
hero member
Activity: 686
Merit: 500
A pumpkin mines 27 hours a night
...as I'm assuming GEN 4 will be available in three/four months....

No way! First of all, gen 3 is viable and will have a longer lifespan than that. But more importantly, friedcat stated that anything below the process of gen 3 will require a lot more work, simulation, etc. He also stated in his April QA that gen 4 will be far more challenging. Even if they're working on it right now, the question will be whether there'll be a tapeout this year. I'd bet on no.

Ahhh ok. This still doesn't change much. I accounted for only selling fifty percent of the chips produced in the next three months, I guess we will just order more if we have too after that and sell them for a cheaper price. I do doubt though that they would fabricate that many chips without an idea for selling at least half within three months of delivery

The key aspect when they can't sell the chips is: self mining / franchising. They get the chips for free practically nothing. Selling the chips is the first priority, as it will yield as much as possible in as little time as possible. Keeping them in stock isn't very wise. So they'll be best off putting them to work for themselves.
Selling half of the chips seems like a reasonably conservative enough guess.
copper member
Activity: 224
Merit: 100
...as I'm assuming GEN 4 will be available in three/four months....

No way! First of all, gen 3 is viable and will have a longer lifespan than that. But more importantly, friedcat stated that anything below the process of gen 3 will require a lot more work, simulation, etc. He also stated in his April QA that gen 4 will be far more challenging. Even if they're working on it right now, the question will be whether there'll be a tapeout this year. I'd bet on no.

Ahhh ok. This still doesn't change much. I accounted for only selling fifty percent of the chips produced in the next three months, I guess we will just order more if we have too after that and sell them for a cheaper price. I do doubt though that they would fabricate that many chips without an idea for selling at least half within three months of delivery
hero member
Activity: 686
Merit: 500
A pumpkin mines 27 hours a night
...as I'm assuming GEN 4 will be available in three/four months....

No way! First of all, gen 3 is viable and will have a longer lifespan than that. But more importantly, friedcat stated that anything below the process of gen 3 will require a lot more work, simulation, etc. He also stated in his April QA that gen 4 will be far more challenging. Even if they're working on it right now, the question will be whether there'll be a tapeout this year. I'd bet on no.
hero member
Activity: 686
Merit: 500
A pumpkin mines 27 hours a night
from 160 000 000 chips total of gen3.
whatt?! where did you get those numbers? Shocked

Its upper limit from :

1) What is the order of magnitude of 3rd generation chips ASICMiner anticipates to sell (total / first batch / first 3 months)?
re 1) Order of magnitude: Depending on fab capacity. Total target is 10k to 40k wafers, relying on competitors, Bitcoin price, and other factors. A wafer is rated at 40 TH/s.

also I see that you're calculating 0.5$ profit per Gh/s ... selling price is from 0.49$ (unlikely for first batches) to 0.99$ (more likely), production costs -0.2$ Gh/s, put this into sheet Wink

so basically, the lower and higher targets takes us between 1 and 4,1 BTC divs per share...   Cheesy Grin Shocked Cool Roll Eyes Tongue Kiss

I remember that discussion. To quote myself back in the days...

On AM value assuming $0.8 per ghs profit; at 200 PH production = 0.6 Btc per share. Anything more, we will be in Btc heaven.

If we're extrapolating: 1600 PH/s production at a profit of $0.80 per GH/s gives approximately XBT 4.8 per share. Not bad, huh?

If we're really looking at $0.8 profit per GH/s, we have to consider that this is only the price for the first month(s). The price for following batches will be lower, since mining will become less profitable. So only the first, say, 50 PH/s will make $0.8 per GH/s, afterwards we have to settle for less. That being said, it's not completely unreasonable to expect another 0.6 BTC per share coming from gen 3. Which would (still) be perfectly fine in my books...

I'm AM bullish, but expecting (a lot) more than 0.5 BTC divs/share for all of gen 3 is delusional these days. Looking at the competition, this is still more than fine. Prepare for 0.25 BTC divs/share and take any surplus as a nice bonus. Disclaimer: This is no trading advice, remember that I may be very wrong about this. Also, I do own shares.
legendary
Activity: 1260
Merit: 1002
from 160 000 000 chips total of gen3.
whatt?! where did you get those numbers? Shocked

Its upper limit from :

1) What is the order of magnitude of 3rd generation chips ASICMiner anticipates to sell (total / first batch / first 3 months)?
re 1) Order of magnitude: Depending on fab capacity. Total target is 10k to 40k wafers, relying on competitors, Bitcoin price, and other factors. A wafer is rated at 40 TH/s.

also I see that you're calculating 0.5$ profit per Gh/s ... selling price is from 0.49$ (unlikely for first batches) to 0.99$ (more likely), production costs -0.2$ Gh/s, put this into sheet Wink

so basically, the lower and higher targets takes us between 1 and 4,1 BTC divs per share...   Cheesy Grin Shocked Cool Roll Eyes Tongue Kiss
hero member
Activity: 560
Merit: 500
Where is the information about the rockminer solo farm?

Somewhere in rockxies ipo thread. IIRC theyhad a goal of selling 10ph.

Looks like hashratio solofarm is up to 115th. Soon to be 1ph:

Is their solo farm consisting purely of AM Gen 3 chips? Potential franchisees?
hero member
Activity: 770
Merit: 509
Quote
OEM service starting order amount: 1,000,000 RMB.

Our OEM price of about 5,000 yuan / sets

About $1/gh.. Not bad if you have 360btc.
member
Activity: 62
Merit: 10
Quote
please God fish testify!

Sounds like a rally cry...
hero member
Activity: 770
Merit: 509
Where is the information about the rockminer solo farm?

Somewhere in rockxies ipo thread. IIRC theyhad a goal of selling 10ph.

Looks like hashratio solofarm is up to 115th. Soon to be 1ph:

Quote
in a few days it should be deployed between the 500T-1000T, please God fish testify!

Side note: I don't think we should expect the first batch of chips to be sold for much more than $0.5/gh considering hashratio manifacturing costs are about $0.8/gh for complete miners.

Quote
Currently we use roasted cat chip production mining machine, stand-alone costs about 4,000 yuan / 768G (overclocked 800G), which is about 5.2RMB / G (without power)

Hashratio info from http://weibo.cn/zhaodong1982
Jump to: