ASICMINER: Entering the Future of ASIC Mining by Inventing It

[freedomno1]

every day for the last 2 weeks this is less true. the real question is how much do you split  your bets

It forever takes two weeks for BFL to ship.

We call it BFL Time
Taken from Valve Time
https://developer.valvesoftware.com/wiki/Valve_Time

HA! so what you are telling me is that BFL will ship the same day as Half Life 3?

Most likely it is so
Heck if Half Life 3 Is the Rapture lets add Two Weeks to BFL

Edit What Saveawedge said

Ok back to thread
The Hardware can make a short term profit but yep depends on duration you want to hold it for
Could still make a profit though if the price goes up to 200 USD Btc again sometime if your willing to mine at a theoretical loss and believe that the exchange rate will appreciate with your mining before the difficulty

[1715074881]

1715074881

[tkone]

if later the bitcoin system can be used to send such small transactions and not be a problem,

why cant we just let it do it now, why is it considered spam? if the system is able to withould this amount a traffic, it shouldnt matter why or who or when these transactions are occured,
it should be allowed with no problem, just take time to confirm thats all...

[helixone]

If you do a "cost analysis" in terms of purchase price vs revenue generated, then:

ASSUMPTIONS
- do not factor in electricity costs
- BTC exchange at today's rate of $120 USD/BTC
- network difficulty on average increases 10% every 2 weeks, thereby reducing revenue by about 10% every 2 weeks

GPU
- AMD 7970 at stock speeds provides 650 MH/s
- Cost $400 USD or 3.33 BTC
- Generates 0.0209 BTC/day at today's difficulty
- After one year of continuous mining total BTC generated = 2.74 BTC
- LOSS OF 0.60 BTC

USB Block Erupter
- 333 MH/s
- Cost 1.99 BTC
- Generates 0.0107 BTC/day at today's difficulty
- After one year of continuous mining total BTC generated = 1.40 BTC
- LOSS OF 0.59 BTC

BFL Jalapeno
- 5000 MH/s
- Cost $274 USD = 2.28 BTC
- Generates 0.1611 BTC/day at today's difficulty
- After one year of continuous mining total BTC generated = 21.1 BTC if you start mining today!
- If you receive your unit and start mining 6 months from now
   - Generates 0.0454 BTC/day
   - After one year from that point, total BTC generated = 5.945 BTC

AM-PT SHARE
- Currently about 2.5 BTC/share at BTC-TC
- Recently generated 0.038 BTC/week in dividends (this was the highest dividend ever)
- After one year of dividends at this rate = 1.98 BTC
- LOSS OF 0.52 BTC

In other words, if you purchase any of the above items in the hopes of "earning back" your cost, you won't (with exception of BFL units). Of course this does not include the fact that you can sell your 7970 GPU in the future (but a year from now how much would it be worth? maybe $100?). Others would say you cannot sell the USB Block Erupter after a year because it is a uni-tasker without any other purpose. The BFL units are reasonably priced. Problem is they are not shipping in bulk, so a delay in shipment leads to large loss in profit. Finally, you can sell your AM-PT share too, and it is very difficult to predict what the sell price would be at that time.

 wrong a hd7790 cost 1btc generates 300MHs  

bfl jallys have sped delivery a lot over the last 10 days  and if you order today and get one in 90 days as claimed your analysis is wrong.

frankly if you want to start in this game today there is no good item to buy.

What happens if you update your model for recent difficulty trends.

IE: Change it from 10% increases every 2 weeks, to 33.6% every two weeks. (The real rate of difficulty increases is over triple what you used, and is likely to only accelerate as other asic hw vendors start shipping gear in volume.)

Also, please remember that a share of AM represents equity in a mining and HW manufacturing company. Unlike HW, and other fixed mining assets, AM shares will not quickly depreciate in value. So at the end of the year, you'll likely be able to sell AM for at least what you paid for (or somewhere close). Good luck trying to sell any of that mining HW in a year for anything near the price you paid for it.

-helixone

[philipma1957]

If you do a "cost analysis" in terms of purchase price vs revenue generated, then:

ASSUMPTIONS
- do not factor in electricity costs
- BTC exchange at today's rate of $120 USD/BTC
- network difficulty on average increases 10% every 2 weeks, thereby reducing revenue by about 10% every 2 weeks

GPU
- AMD 7970 at stock speeds provides 650 MH/s
- Cost $400 USD or 3.33 BTC
- Generates 0.0209 BTC/day at today's difficulty
- After one year of continuous mining total BTC generated = 2.74 BTC
- LOSS OF 0.60 BTC

USB Block Erupter
- 333 MH/s
- Cost 1.99 BTC
- Generates 0.0107 BTC/day at today's difficulty
- After one year of continuous mining total BTC generated = 1.40 BTC
- LOSS OF 0.59 BTC

BFL Jalapeno
- 5000 MH/s
- Cost $274 USD = 2.28 BTC
- Generates 0.1611 BTC/day at today's difficulty
- After one year of continuous mining total BTC generated = 21.1 BTC if you start mining today!
- If you receive your unit and start mining 6 months from now
   - Generates 0.0454 BTC/day
   - After one year from that point, total BTC generated = 5.945 BTC

AM-PT SHARE
- Currently about 2.5 BTC/share at BTC-TC
- Recently generated 0.038 BTC/week in dividends (this was the highest dividend ever)
- After one year of dividends at this rate = 1.98 BTC
- LOSS OF 0.52 BTC

In other words, if you purchase any of the above items in the hopes of "earning back" your cost, you won't (with exception of BFL units). Of course this does not include the fact that you can sell your 7970 GPU in the future (but a year from now how much would it be worth? maybe $100?). Others would say you cannot sell the USB Block Erupter after a year because it is a uni-tasker without any other purpose. The BFL units are reasonably priced. Problem is they are not shipping in bulk, so a delay in shipment leads to large loss in profit. Finally, you can sell your AM-PT share too, and it is very difficult to predict what the sell price would be at that time.

 wrong a hd7790 cost 1btc generates 300MHs  

bfl jallys have sped delivery a lot over the last 10 days  and if you order today and get one in 90 days as claimed your analysis is wrong.

frankly if you want to start in this game today there is no good item to buy.

What happens if you update your model for recent difficulty trends.

IE: Change it from 10% increases every 2 weeks, to 33.6% every two weeks. (The real rate of difficulty increases is over triple what you used, and is likely to only accelerate as other asic hw vendors start shipping gear in volume.)

Also, please remember that a share of AM represents equity in a mining and HW manufacturing company. Unlike HW, and other fixed mining assets, AM shares will not quickly depreciate in value. So at the end of the year, you'll likely be able to sell AM for at least what you paid for (or somewhere close). Good luck trying to sell any of that mining HW in a year for anything near the price you paid for it.

-helixone

Frankly  sitting on the sidelines and watching the show may be better then any investment at all.   I can give 5 bad points to every BTC investment today.  Your point of big hash ramp up is true.  we may go from 15 to 18 to 22 then 26     20% each jump up.  This is very likely to happen .

If true and asicminer ramps right up with it then AM will be fine. 

 So on July 1 when the network is 220Th if AM is 66TH  AM will be okay even if AM can't sell its usb sticks any more.

Or has dropped price to .5BTC for a stick and 8BTC for a Blade  of course I am saying those numbers because they would be the correct price if BFL has finally turned the corner.  Or some other developer uses a bunch of avalon chips to make a good machine at a good price.  This summer will have lots of action.   

One thing for sure gpus will end hashing  for BTC in 2013.

[MikkisJ]

I tried myself to create a list with AM shareholders based on yesterday's dividends:

https://docs.google.com/spreadsheet/ccc?key=0AtqphFCP56ordGVCakJxSU90MlB4MlBkZENya25pS2c

I got > 398k of the 400k shares. If you find any errors please let me know!

Since there are almost 400,000 shares on the list it mean PT shares are included. But there are no transfers with around 20,000 shares (btct.co and bitfunder). So it must mean the pass through exchanges don't get dividends in one transfer, but multiple.

Which of the transfers belong to pass throughs?

[crashoveride54902]

I tried myself to create a list with AM shareholders based on yesterday's dividends:

https://docs.google.com/spreadsheet/ccc?key=0AtqphFCP56ordGVCakJxSU90MlB4MlBkZENya25pS2c

I got > 398k of the 400k shares. If you find any errors please let me know!

dayum!! I'd sure like to have any address above line 200 instead i think my addy is the last one oh well i'll take it and just try and hold my envy for all of you here...

Whoever says AM is a bad investment, I hope they are wrong I decided to put all my mined btc into AM shares as i saw the end of my mining quickly approaching...and still wanted to have btc income w/o having to buy it or mine at a loss...so i was glad when i found AM and btc-tc...granted it won't be huge income...but still better then none

[stripykitteh]

  MY concern is over the usb sticks if 4000 have been sold for 8000 btc do we have 6000 left to sell?  What will we try to move them at 2btc A STICK.? MY next concern is how many blade have been sold? How many are left to sell and will we still try to sell them at 30 BTC a blade.  Seems to me all the sales will drop a lot and dividends will be greatly reduced very soon as BFL will have all jalapenos delivered and up to real time in 90 days. see links

http://www.coindesk.com/butterfly-labs-finally-ships-out-last-years-jalapeno-orders/

https://forums.butterflylabs.com/blogs/bfl_jody/180-wednesday-6-5-2012.html

 IF true a BFL 5gh sold for 2.5btc that  will end AM income from sales.  This means AM will need more hash and less sales or far lower prices to sell enough  To keep current dividends high.  


Or we can all hope BFL will stop shipping the jalapenos at a fast rate.

We have heard a lot about BFL shipping 5GH/s devices. The problem BFL have is that is all they can ship. It appears that most of the chips used in the currently shipping devices are doing ~2.7GH/s each (2 chips/device). If you want the 7GH/s device, they want to charge you an extra $100 to use 2 ~3.7GH/s chips. The inference I draw from this is that their 'best' chips are ~3.7GH/s, but they have many more which are lower performing or do not work at all.

Their 25GH/s and 50GH/s devices will require chips with a minimum of 3.125GH/s hashrate each (8 for the little and 16 for the big) to meet the specs. That requires that BFL have those chips in volume. The next inference I draw is that if they chose to ship those bigger devices they'd go broke, which is why they're not shipping them.

The order book for big devices is a huge liability for BFL.

[Franktank]

I tried myself to create a list with AM shareholders based on yesterday's dividends:

https://docs.google.com/spreadsheet/ccc?key=0AtqphFCP56ordGVCakJxSU90MlB4MlBkZENya25pS2c

I got > 398k of the 400k shares. If you find any errors please let me know!

Since there are almost 400,000 shares on the list it mean PT shares are included. But there are no transfers with around 20,000 shares (btct.co and bitfunder). So it must mean the pass through exchanges don't get dividends in one transfer, but multiple.

Which of the transfers belong to pass throughs?

Hint: On the passthrus, you can see how many each hold (BF/BTCT). Check out the G Doc listed above to see which address that holds that number of shares. Presto...

[burnside]

I tried myself to create a list with AM shareholders based on yesterday's dividends:

https://docs.google.com/spreadsheet/ccc?key=0AtqphFCP56ordGVCakJxSU90MlB4MlBkZENya25pS2c

I got > 398k of the 400k shares. If you find any errors please let me know!

Since there are almost 400,000 shares on the list it mean PT shares are included. But there are no transfers with around 20,000 shares (btct.co and bitfunder). So it must mean the pass through exchanges don't get dividends in one transfer, but multiple.

Which of the transfers belong to pass throughs?

I quoted the txid of our last div payment in the BTCT.CO ASICMINER-PT thread.  It all came as one transaction.

[MikkisJ]

I tried myself to create a list with AM shareholders based on yesterday's dividends:

https://docs.google.com/spreadsheet/ccc?key=0AtqphFCP56ordGVCakJxSU90MlB4MlBkZENya25pS2c

I got > 398k of the 400k shares. If you find any errors please let me know!

Since there are almost 400,000 shares on the list it mean PT shares are included. But there are no transfers with around 20,000 shares (btct.co and bitfunder). So it must mean the pass through exchanges don't get dividends in one transfer, but multiple.

Which of the transfers belong to pass throughs?

Hint: On the passthrus, you can see how many each hold (BF/BTCT). Check out the G Doc listed above to see which address that holds that number of shares. Presto...

How does it work? Does AM make one big payment to pass throughs and the pass through makes small payments inside? Or AM pays dividends to each share holder inside of a pass through?

Edit, while typing an answer appeared by burnside. Does that work the same way with BitFunder?

[GuiltySpark343]

wrong a hd7790 cost 1btc generates 300MHs  

WTF? I didn't even mention a 7790, so how can I be wrong?

[boyohi]

I tried myself to create a list with AM shareholders based on yesterday's dividends:

https://docs.google.com/spreadsheet/ccc?key=0AtqphFCP56ordGVCakJxSU90MlB4MlBkZENya25pS2c

I got > 398k of the 400k shares. If you find any errors please let me know!

Since there are almost 400,000 shares on the list it mean PT shares are included. But there are no transfers with around 20,000 shares (btct.co and bitfunder). So it must mean the pass through exchanges don't get dividends in one transfer, but multiple.

Which of the transfers belong to pass throughs?

Hint: On the passthrus, you can see how many each hold (BF/BTCT). Check out the G Doc listed above to see which address that holds that number of shares. Presto...

How does it work? Does AM make one big payment to pass throughs and the pass through makes small payments inside? Or AM pays dividends to each share holder inside of a pass through?

Edit, while typing an answer appeared by burnside. Does that work the same way with BitFunder?

AM sends the dividends to burnside or deadterra and those people sends it out to that shares in the PT. Its indirect and you never come into contact with friedcat.

[GuiltySpark343]

The comparison is faulty because AM is a company and the other alternatives are hardware.

Why is that relevant, you ask? Well, AM will (hopefully) continuously renew itself and its strategy in order to keep earnings high, whereas the hardware simply gets outdated and earns progressively less and less as time goes along. So: buying hardware may earn you more in the short term, shares in AM may earn you more in the long term.

Comparison is NOT faulty. It is a comparison of "investment" (whether in hardware or a company) and "return" (whether mined bitcoins or dividends paid out as bitcoins). I think most people will agree that the value of hardware (GPU or ASICs) will generally depreciate over time, so that is a "known." The big unknown is AM, which as you point out is a company. We hope they will continuously renew itself, but anything can happen.

[philipma1957]

  MY concern is over the usb sticks if 4000 have been sold for 8000 btc do we have 6000 left to sell?  What will we try to move them at 2btc A STICK.? MY next concern is how many blade have been sold? How many are left to sell and will we still try to sell them at 30 BTC a blade.  Seems to me all the sales will drop a lot and dividends will be greatly reduced very soon as BFL will have all jalapenos delivered and up to real time in 90 days. see links

http://www.coindesk.com/butterfly-labs-finally-ships-out-last-years-jalapeno-orders/

https://forums.butterflylabs.com/blogs/bfl_jody/180-wednesday-6-5-2012.html

 IF true a BFL 5gh sold for 2.5btc that  will end AM income from sales.  This means AM will need more hash and less sales or far lower prices to sell enough  To keep current dividends high.  


Or we can all hope BFL will stop shipping the jalapenos at a fast rate.

We have heard a lot about BFL shipping 5GH/s devices. The problem BFL have is that is all they can ship. It appears that most of the chips used in the currently shipping devices are doing ~2.7GH/s each (2 chips/device). If you want the 7GH/s device, they want to charge you an extra $100 to use 2 ~3.7GH/s chips. The inference I draw from this is that their 'best' chips are ~3.7GH/s, but they have many more which are lower performing or do not work at all.

Their 25GH/s and 50GH/s devices will require chips with a minimum of 3.125GH/s hashrate each (8 for the little and 16 for the big) to meet the specs. That requires that BFL have those chips in volume. The next inference I draw is that if they chose to ship those bigger devices they'd go broke, which is why they're not shipping them.

The order book for big devices is a huge liability for BFL.

 

   yep and they promise to start shipping them soon. (2 weeks)  but if they offer  multiple 5ghs units to replace the little single and the single people will take them.

   I don't consider them to be a great company and I cancelled all my orders for little singles months ago  use that money to invest in AM made some good coin.  I have no complaints about AM from march 2013  to now.  they did nice by money.  I am pointing out that a 5GH/s unit for 2.5 coins is a lot cheaper then a usb stick for 2 coins  since you need 16 usb sticks to hash at the rate of 1 bfl unit.  I was pretty convinced that BFL would fail and I did cancel my larger orders in March  And in early May.  


 Once again the BFL  5GH unit is about 2GH per btc.

 The  AM 334MH unit is about 167MH per btc.   The AM blade is about 200MH per btc.   These units will stop selling very soon at those prices.  If BFL continues to ship. (big if)

  While a good amount of usb sticks sold 4000 there are 6000 not sold.  the 2BTC was always too high of a price.  When it first came out I pointed this out by comparing it to hd7790's my point was simple anyone with an empty slot in a pc was better off using hd7790 as it was 1btc for 300 hash and the stick was 333 hash for 2 BTC.  the 1 coin price difference takes more then 200 days to make up for in power savings.   A hd7790 cost 2 kwatt a day if your pc has room for it. at .25 usd a kwatt that would be .50 usd a day for power that means 240 days to be the 1 btc price difference. I purchased 8 hd7790's on APRIL  1 they hashed at 2400MH and earned  over 8btc back minus power about  7btc.  I was lucky and the first few coins earned  were high priced so my 8 cards now cost about 1btc.

  if I purchased  8 usb sticks my cost would have been 16btc I would be down  9btc.   this window is dying now since the power edge for the usb sticks will soon make them better then a hd7790.  the problem is the bfl 5gh unit crushes a usb stick and 2 bfl 5gh units bitch slap a blade. So the sales window for AM is closing at the current prices they charge.  Since no one knows ( a few do not us) what a stick or a blade really cost none of us know how low AM can price them.   So As I have been saying  AM's sales are going to drop in price or volume which means AM has to maintain its mining percentage to stay on top.  Hell for all I know AM can go to 150Th as I type and they are just going to keep at that 20-30 % level for years to come. I just do not see the equipment sales continuing to whale any more. Thus dividends will be lower. more like .02 not .038 a share. I see AM at 1.5BTC a share by Aug 1.  

But hell this is a market not real math and I am missing tons of info.  

[philipma1957]

wrong a hd7790 cost 1btc generates 300MHs  

WTF? I didn't even mention a 7790, so how can I be wrong?

 right leaving out a hd7790 and using a hd7970 is wrong since a 7790 is the cheapest gpu for btc mining based on hash /power / price.  


 A hd7790 is 300 hash for 80 watts at 1.1 btc  

 a hd7970 is 650 hash for 240 watts at 3.3 btc    


 at todays rates  the hd7970 makes about $1.55 a day   2 hd7790's make about $1.67 a day    and you spent 3 btc to get the hd7970 and 2 btc to get 2 hd7790's...  of course if you are short slots some will argue the hd7970 is better. in my case I sold off used hd7970's and replaced with hd7790's worked out well sell one used hd7970 and get 3 new hd7790's pretty good deal.

[radiumsoup]

The comparison is faulty because AM is a company and the other alternatives are hardware.

Why is that relevant, you ask? Well, AM will (hopefully) continuously renew itself and its strategy in order to keep earnings high, whereas the hardware simply gets outdated and earns progressively less and less as time goes along. So: buying hardware may earn you more in the short term, shares in AM may earn you more in the long term.

Comparison is NOT faulty. It is a comparison of "investment" (whether in hardware or a company) and "return" (whether mined bitcoins or dividends paid out as bitcoins). I think most people will agree that the value of hardware (GPU or ASICs) will generally depreciate over time, so that is a "known." The big unknown is AM, which as you point out is a company. We hope they will continuously renew itself, but anything can happen.

don't mean to burst your bubble, but yeah, it IS faulty. Everything in that list has a decelerating profit vector due to the fact that the hardware cannot be upgraded and has a definite lifespan (assuming hash rate increases over time - a pretty safe bet). The exception is AM, which not only mines using hardware like the other 3 examples, but also sells hardware to others, effectively offloading the lifespan-based risk at a discount and keeping the instant profits, which are then redistributed to shareholders.

It's rather like comparing 3 fixed-rate CDs versus a growth stock during a time of inflation. The CDs will perform less and less over time, but the growth stock is not limited to the fixed return that the CDs are, even though both are affected the same by inflation. (inflation is analogous to hash rate in this example)

[bitfair]

The comparison is faulty because AM is a company and the other alternatives are hardware.

Why is that relevant, you ask? Well, AM will (hopefully) continuously renew itself and its strategy in order to keep earnings high, whereas the hardware simply gets outdated and earns progressively less and less as time goes along. So: buying hardware may earn you more in the short term, shares in AM may earn you more in the long term.

Comparison is NOT faulty. It is a comparison of "investment" (whether in hardware or a company) and "return" (whether mined bitcoins or dividends paid out as bitcoins). I think most people will agree that the value of hardware (GPU or ASICs) will generally depreciate over time, so that is a "known." The big unknown is AM, which as you point out is a company. We hope they will continuously renew itself, but anything can happen.

First, let me apologize to all the other readers of this thread for harping on about this.

Secondly, I guess "value" is a very subjective notion and I am perfectly happy to agree to disagree on this.

Furthermore, I agree with you that the future is uncertain - also with respect to AMs future performance, it is by no means guaranteed that it will continue to be successful.

But the difference between AM and a piece of hardware is that AM is capable of strategic adaptation. I concede that a specific piece of hardware may outperform AM for a short period (your example arbitrarily showed one year, and I suppose that is OK for your purposes, but my horizon is longer). Beyond the very short term, however, the capability to adapt becomes more important: in three years the piece of hardware you purchased today will continue to hash at 5 GH/s regardless of what the network hash rate is (i.e. essentially earning very low returns), but at that time I expect AM has expanded and evolved way beyond what it is today and continues to rake in large profits.

It is uncertain whether or not AM will still be raking in huge profits in three years. But I can guarantee that your 5 GH/s miner will be more or less worthless at that time.

Those are just my thoughts, and anybody is of course welcome to disagree and invest as they see fit, focusing on the time horizon or risk profile that fits their personal taste. So now that I shared that, I will stop harping on about it - back to your regularly scheduled programme...

[philipma1957]

The comparison is faulty because AM is a company and the other alternatives are hardware.

Why is that relevant, you ask? Well, AM will (hopefully) continuously renew itself and its strategy in order to keep earnings high, whereas the hardware simply gets outdated and earns progressively less and less as time goes along. So: buying hardware may earn you more in the short term, shares in AM may earn you more in the long term.

Comparison is NOT faulty. It is a comparison of "investment" (whether in hardware or a company) and "return" (whether mined bitcoins or dividends paid out as bitcoins). I think most people will agree that the value of hardware (GPU or ASICs) will generally depreciate over time, so that is a "known." The big unknown is AM, which as you point out is a company. We hope they will continuously renew itself, but anything can happen.

...

But the difference between AM and a piece of hardware is that AM is capable of strategic adaptation...

 I completely agree that AM can bust a move and when they bust one I will most likely buy in again.

   For now I see them with 6000 usb sticks  not sold and no new pricing strategy.   Now I concede I am not an insider and I have 0 info on what BFL can sell in the future but they have finished all of JULY 2012 and have started on August 2012.  On May 23 they had  June 23-25  shipped. So 2 weeks later (boy do we love that 2 weeks) they are up to August of 2012.  Just remember  1x  5ghs unit at 2.5 BTC = 16 sticks at a total of 32 BTC .  AM selling sticks  at 2 BTC may have looked good to most stock owners of AM but I always stated they were overpriced. Go back and look at my posts. So to me it looks like short term thinking on the part of AM. I am not second guessing since I said this at the get go.  

  Still the sticks may not be a big problem all that matters is AM paces the hash percentage correctly 22-32% and AM does not need to sell a thing just mine away and let others do the selling of the picks and shovels so to speak.

[daemondazz]

I'm going to hazard a guess that the USB miners probably cost around $USD30-40 to make, which at $120/BTC would make them around 0.25BTC. There is plenty of room for them to drop the price (and quickly) as and when some actual competition appears.

[Eric Muyser]

I'm going to hazard a guess that the USB miners probably cost around $USD30-40 to make, which at $120/BTC would make them around 0.25BTC. There is plenty of room for them to drop the price (and quickly) as and when some actual competition appears.

This ^ AM is smart.

They are capitalizing on their investment and IP.

Intel does it too. They are so mainstream and have AMD to compete with though that the generations progress and prices decrease at a higher pace. AM has no such AMD, just a bunch of wannabe's as of yet.