Pages:
Author

Topic: [ANN]ASICMiner Publicly Looking for Potential Customers/Partners for New Chips - page 2. (Read 54971 times)

legendary
Activity: 1666
Merit: 1183
dogiecoin.com
There's a sucker born every minute, your going to have to find them to buy this under performing tech.
will probably cost less than $2.00 to make

Less.
sr. member
Activity: 441
Merit: 250
Why bother re-inventing the wheel, spending probably $3m to get someone to do it for you and missing probably 6-9 months of earnings? It's a no brainer, even with an inefficient solution. With the Bitcoin price rebounding  (for now) it looks more and more attractive.....
legendary
Activity: 1904
Merit: 1007
There's a sucker born every minute, your going to have to find them to buy this under performing tech.
I think you may be underestimating the power of the locomotive that is driving this program. Based on what I know about the chip, even if it is a bit of a 'dog' in terms of power consumption, it's small enough to yield in the 90's of percent and will probably cost less than $2.00 to make, test and package in high volume. A 4TH system using 512 of them -running in near threshold mode - could probably be built at a cost of around $1700, including the PSU's. If 'someone' had access to about $15m in liquid capital, they could procure about 6000 of these machines - 3 million chips, or 50 wafer batches, rent a big warehouse in Washington state, pay a few million in setup costs and installation of a 50MVA substation then that could easily be 24PH on hand within 6 months from a standing start (which may have started 2 months ago); with cheap electricity at less than 4 cents/kWh, power consumption becomes a lot less important.

Let's say that difficulty in October is 30 x 10e9 and BTC at $450. This installation would earn about $4.5 million in it's first month, even if the rigs and their cooling gobble up power at 1.2 Joules/gigahash. The 'suckers' would break even in just over 6 months allowing for running costs, but then they might decide to buy some more capacity - they can afford to and there's plenty spare capacity in that substation (and on TSMC's 40nm lines). Once hashing capacity is built and installed and has paid back it's manufacturing, installation costs and interest it can keep on hashing until the power and running costs get too high, and that might take a year or two - think an ROI of over 500%.

Individual miners do not have these resources, even the ones who have 'free' power (whatever that is, it won't last for long once their landlords start realising why their energy bills are so high).

Summary: don't underestimate your competition.



Money don't grow on trees! Wake up!

I think you're fooling yourself if you truly believe that. However, that''s your perogative.

Someone with 15m$ in hand should be better designing their own chip than to buy the underperforming AM chip anyway.
full member
Activity: 161
Merit: 100
There's a sucker born every minute, your going to have to find them to buy this under performing tech.
I think you may be underestimating the power of the locomotive that is driving this program. Based on what I know about the chip, even if it is a bit of a 'dog' in terms of power consumption, it's small enough to yield in the 90's of percent and will probably cost less than $2.00 to make, test and package in high volume. A 4TH system using 512 of them -running in near threshold mode - could probably be built at a cost of around $1700, including the PSU's. If 'someone' had access to about $15m in liquid capital, they could procure about 6000 of these machines - 3 million chips, or 50 wafer batches, rent a big warehouse in Washington state, pay a few million in setup costs and installation of a 50MVA substation then that could easily be 24PH on hand within 6 months from a standing start (which may have started 2 months ago); with cheap electricity at less than 4 cents/kWh, power consumption becomes a lot less important.

Let's say that difficulty in October is 30 x 10e9 and BTC at $450. This installation would earn about $4.5 million in it's first month, even if the rigs and their cooling gobble up power at 1.2 Joules/gigahash. The 'suckers' would break even in just over 6 months allowing for running costs, but then they might decide to buy some more capacity - they can afford to and there's plenty spare capacity in that substation (and on TSMC's 40nm lines). Once hashing capacity is built and installed and has paid back it's manufacturing, installation costs and interest it can keep on hashing until the power and running costs get too high, and that might take a year or two - think an ROI of over 500%.

Individual miners do not have these resources, even the ones who have 'free' power (whatever that is, it won't last for long once their landlords start realising why their energy bills are so high).

Summary: don't underestimate your competition.



Money don't grow on trees! Wake up!

I think you're fooling yourself if you truly believe that. However, that''s your perogative.
legendary
Activity: 1904
Merit: 1007
There's a sucker born every minute, your going to have to find them to buy this under performing tech.
I think you may be underestimating the power of the locomotive that is driving this program. Based on what I know about the chip, even if it is a bit of a 'dog' in terms of power consumption, it's small enough to yield in the 90's of percent and will probably cost less than $2.00 to make, test and package in high volume. A 4TH system using 512 of them -running in near threshold mode - could probably be built at a cost of around $1700, including the PSU's. If 'someone' had access to about $15m in liquid capital, they could procure about 6000 of these machines - 3 million chips, or 50 wafer batches, rent a big warehouse in Washington state, pay a few million in setup costs and installation of a 50MVA substation then that could easily be 24PH on hand within 6 months from a standing start (which may have started 2 months ago); with cheap electricity at less than 4 cents/kWh, power consumption becomes a lot less important.

Let's say that difficulty in October is 30 x 10e9 and BTC at $450. This installation would earn about $4.5 million in it's first month, even if the rigs and their cooling gobble up power at 1.2 Joules/gigahash. The 'suckers' would break even in just over 6 months allowing for running costs, but then they might decide to buy some more capacity - they can afford to and there's plenty spare capacity in that substation (and on TSMC's 40nm lines). Once hashing capacity is built and installed and has paid back it's manufacturing, installation costs and interest it can keep on hashing until the power and running costs get too high, and that might take a year or two - think an ROI of over 500%.

Individual miners do not have these resources, even the ones who have 'free' power (whatever that is, it won't last for long once their landlords start realising why their energy bills are so high).

Summary: don't underestimate your competition.



Money don't grow on trees! Wake up!
full member
Activity: 161
Merit: 100
There's a sucker born every minute, your going to have to find them to buy this under performing tech.
I think you may be underestimating the power of the locomotive that is driving this program. Based on what I know about the chip, even if it is a bit of a 'dog' in terms of power consumption, it's small enough to yield in the 90's of percent and will probably cost less than $2.00 to make, test and package in high volume. A 4TH system using 512 of them -running in near threshold mode - could probably be built at a cost of around $1700, including the PSU's. If 'someone' had access to about $15m in liquid capital, they could procure about 6000 of these machines - 3 million chips, or 50 wafer batches, rent a big warehouse in Washington state, pay a few million in setup costs and installation of a 50MVA substation then that could easily be 24PH on hand within 6 months from a standing start (which may have started 2 months ago); with cheap electricity at less than 4 cents/kWh, power consumption becomes a lot less important.

Let's say that difficulty in October is 30 x 10e9 and BTC at $450. This installation would earn about $4.5 million in it's first month, even if the rigs and their cooling gobble up power at 1.2 Joules/gigahash. The 'suckers' would break even in just over 6 months allowing for running costs, but then they might decide to buy some more capacity - they can afford to and there's plenty spare capacity in that substation (and on TSMC's 40nm lines). Once hashing capacity is built and installed and has paid back it's manufacturing, installation costs and interest it can keep on hashing until the power and running costs get too high, and that might take a year or two - think an ROI of over 500%.

Individual miners do not have these resources, even the ones who have 'free' power (whatever that is, it won't last for long once their landlords start realising why their energy bills are so high).

Summary: don't underestimate your competition.

legendary
Activity: 2576
Merit: 1186
Some questions...

What is the behaviour if task address is changed while the chip is busy hashing?

What about when address 44 is changed?

How is nonce_mask to be interpreted?

Is there any safeguard against a race clearing r_ready after reading nonce(s)?
For example, the order of events:
  • Host reads nonce from chip
  • Chip finds new nonce
  • Host clears r_ready
sr. member
Activity: 434
Merit: 250
Keep in mind that those numbers where agreed on following the original chip specs.

However, currently delivered chips get a maximum of around 0,8W/GH in completed devices, so the actual order amount could be considerably lower, as the chips are way off-spec.

I don't think any chips were sold before AM had them in hand and tested.

still waiting on dividends?  Undecided
hero member
Activity: 728
Merit: 500
There's a sucker born every minute, your going to have to find them to buy this under performing tech.
full member
Activity: 224
Merit: 100
These numbers are derived likely running them near max capacity while unit costs are high. AM should have a long term price advantage to get more ASICs into systems when the economics for this requirement come into play. Unfortunately we haven't the full operating range data yet to see what opportunities exist in that regard.
sr. member
Activity: 434
Merit: 250
Keep in mind that those numbers where agreed on following the original chip specs.

However, currently delivered chips get a maximum of around 0,8W/GH in completed devices, so the actual order amount could be considerably lower, as the chips are way off-spec.

I don't think any chips were sold before AM had them in hand and tested.

https://bitcointalksearch.org/topic/btcgarden-am-v1-down-to-053ghs-in-stock-for-international-selling-622439

No AM gen3 complete device showed any number below 1.1 W/GHs
hero member
Activity: 770
Merit: 509
Keep in mind that those numbers where agreed on following the original chip specs.

However, currently delivered chips get a maximum of around 0,8W/GH in completed devices, so the actual order amount could be considerably lower, as the chips are way off-spec.

I don't think any chips were sold before AM had them in hand and tested.
hero member
Activity: 714
Merit: 500
Keep in mind that those numbers where agreed on following the original chip specs.

However, currently delivered chips get a maximum of around 0,8W/GH in completed devices, so the actual order amount could be considerably lower, as the chips are way off-spec.
legendary
Activity: 1260
Merit: 1002
how many guys here have received the first batch of AM 3rd generation chips ? It seems so quite here and  just feel nothing happened still...

yea i believe there was some sort of non disclosure agreement between FC and buyers so that until they launch their products, no one has a clue on FC's order book depth and sales.. but bear in mind that batch 1, 2 & 3 combined are supposedly > 10 Million chips.  Shocked  Cheesy
full member
Activity: 134
Merit: 100
how many guys here have received the first batch of AM 3rd generation chips ? It seems so quite here and  just feel nothing happened still...
full member
Activity: 224
Merit: 100
I often make a joke to my wife about my cooker mining bitcoins... once you switch off the oven, the oven fan continues to run to dispel the excess heat, it sounds like a little mining rig... but in fact, it's not too far from possible, stick in a wireless transmitter, embed the f/w and there you have it, roast beef and bitcoins.  Cheesy

Miners should come with a Teflon top so they're dual purpose. Grin They kept my basement fireplace from coming all winter. Have to find something useful for them to do with the heat in summer. If I can figure out a way for them to heat the pool I've got it covered for all seasons.
sr. member
Activity: 434
Merit: 250
What if:

1. Netgear / Cisco / DLink / Zyxel / TPLink/ Asus / Belkin / a bunch of router / switch manufacturers imbed a chip in their hardware, and lower the price of their products, so more people buy them. Most people leave their routers on, all the time.
This will be the norm in the future.  Anything that has an internet connection and power will have a swappable card that has a Bitcoin ASIC on it.  Btw, that appliance/hardware may be close to "free" if you let the manufacturer mine.

They just better make sure that no one hacks the hardware to mine to a different pool or a different address. Happens all the time with console boxes (Sony Playstation, Microsoft X-Box, etc.).

I'm in for a free router. Why not? As long as it's still 10 watts or something.

I often make a joke to my wife about my cooker mining bitcoins... once you switch off the oven, the oven fan continues to run to dispel the excess heat, it sounds like a little mining rig... but in fact, it's not too far from possible, stick in a wireless transmitter, embed the f/w and there you have it, roast beef and bitcoins.  Cheesy
hero member
Activity: 489
Merit: 500
Immersionist
Don't most people get their routers from the ISPs (most as in the majority of all people that use internet, not the majority of geeky people)
legendary
Activity: 3416
Merit: 1912
The Concierge of Crypto
What if:

1. Netgear / Cisco / DLink / Zyxel / TPLink/ Asus / Belkin / a bunch of router / switch manufacturers imbed a chip in their hardware, and lower the price of their products, so more people buy them. Most people leave their routers on, all the time.
This will be the norm in the future.  Anything that has an internet connection and power will have a swappable card that has a Bitcoin ASIC on it.  Btw, that appliance/hardware may be close to "free" if you let the manufacturer mine.

They just better make sure that no one hacks the hardware to mine to a different pool or a different address. Happens all the time with console boxes (Sony Playstation, Microsoft X-Box, etc.).

I'm in for a free router. Why not? As long as it's still 10 watts or something.
donator
Activity: 2352
Merit: 1060
between a rock and a block!
What if:

1. Netgear / Cisco / DLink / Zyxel / TPLink/ Asus / Belkin / a bunch of router / switch manufacturers imbed a chip in their hardware, and lower the price of their products, so more people buy them. Most people leave their routers on, all the time.
This will be the norm in the future.  Anything that has an internet connection and power will have a swappable card that has a Bitcoin ASIC on it.  Btw, that appliance/hardware may be close to "free" if you let the manufacturer mine.
Pages:
Jump to: