[Announcement] Avalon ASIC Development Status [Batch #1]

[Inaba]

Hey Ice_chill,

Hard for Tom to be friendly with BFL when Josh trolls his thread mercilessly.

Except that, I only post in Tom's threads when he posts lies about myself or BFL.  So... it's not really trolling if it's responding to posts Tom has made.  If Tom doesn't want me to post in his threads, he should stop lying.  Nice try though.

[1713310341]

1713310341

[1713310341]

1713310341

[BitSyncom]

With such high power usage, it seems you'd need to at least double the hash rate to be competitive from a pricing standpoint.

Run the numbers then speak to me again.

When designing your hardware, please don't assume that only people who can get power for <= $0.10 would like to "mine for profit". Power usage is a significant concern if your competitors are offering 60GH/s @ 60W.

That's not what I said, nor what we at Avalon did. I made a statement on long term mining as an investment, stating if you can not get access to cheap power, do not go into mining.

this would be true only in the case that the difficulty will remain unchanged for 5 months, and you know that it will not

real ROI will be, at best, somewhere at 8 months.

8 month based on what? I would much prefer if you have some numbers to back this claim up.

In addition, in case you missed my point. These numbers are made based on the assumption ASIC miners are mining at this 6x or 10x difficulty from day one, this will not happen at all. This is similar to saying the difficulty will raise 6x or 10x before any ASICs are delivered. which as you know, clearly is not the case.

In fact, by stating this 10x estimate, I am saying I predict within 6 month, the network speed will go up 10x, even when by the time all the ASICs become online, and difficulty goes up your ROI to start mining at 6 month later will be 156 days I gave out, which only means one thing, that your ROI will definitely be much less.

I have already gave a really conservative estimate on ROI, unless anybody can show me some other numbers that reach your conclusion of how Power Consumption matters a lot in ROI, I would be much appreciated if people do not speak with such prejudice about power consumption and spread FUD in this thread.

Thank you.

Good news! Do you have any idea when the non-preorder, retail ordering will start?

Very soon now, we still have more exciting news to release, and after we tackle a potential shipping delay around Chinese New Year. Once we mitigated this issue we can offer a better picture on when these will ship out. Unlike our competition, we will release much smaller batches every week or so, and they will ship out within the week. Meaning our customers will always receive their product quickly as possible and we do not get stuck with back orders which we have to fill.

[Inaba]

The difficulty will be 10x within four weeks of the first ASICs that start shipping, be it BFL or Tom's offering.  After a month, most of the back orders of both companies will be well into consumers hands and the difficulty will rise.  6 months?  It will be a lot more than 10x by then.

Your ROI vs Power Consumption is based on your cost for power and if you're assuming cheap power to base your calculations on, I think you're doing a disservice to a lot of miners.  There's no reason to only mine if you have cheap power - you should buy power efficient devices. 

It's pretty irresponsible to even suggest that power doesn't matter, if for no other reason than the environmental impact.  It's a travesty when people mine with inefficient devices just because they aren't the ones having to pay for that power.  But someone, somewhere, is having to pay and it's costing the environment no matter who's paying the monetary bill. Everyone should be striving to have the most power efficient mining devices out there.

[Tinua]

What do you think about this?
Especially as regards the European market, with electricity costs between $ 0.10-0.30/kwh

strictly speaking on power being the only variable I wouldn't even suggest to "mine-for-profit" basis with power rates higher than $0.10 ( and you are unable to get write offs or reductions. )

Sorry, but you know, in how many countries in the world the power rates is cheaper than $0.10, right?
In Europe I know none! This means, that you throw away in minimum the whole European market for  "mine-for-profit" or "long term mining as an investment" customers!

[PuertoLibre]

The difficulty will be 10x within four weeks of the first ASICs that start shipping, be it BFL or Tom's offering.  After a month, most of the back orders of both companies will be well into consumers hands and the difficulty will rise.  6 months?  It will be a lot more than 10x by then.

Congrats, you just made a point that hurts the ASIC sales across the board.

If your words are taken literally, then the ASIC sellers are selling people very expensive lemons that will take a year or more to get their investment back....and then.....*barely any profit* per month.

Your ROI vs Power Consumption is based on your cost for power and if you're assuming cheap power to base your calculations on, I think you're doing a disservice to a lot of miners.  There's no reason to only mine if you have cheap power - you should buy power efficient devices.  

It's pretty irresponsible to even suggest that power doesn't matter, if for no other reason than the environmental impact.  It's a travesty when people mine with inefficient devices just because they aren't the ones having to pay for that power.  But someone, somewhere, is having to pay and it's costing the environment no matter who's paying the monetary bill. Everyone should be striving to have the most power efficient mining devices out there.

Tethered wattage means your device will consume anywhere from 60watts on up:

Not including a standard router wattage: 20watts~

If a laptop: 30 to 65 additional watts.

If a barebones desktop: Anywhere from 80 to 160 watts. (not including the monitor)

        If a fully loaded desktop with 1 or more GPU cards....I care not to speculate. (These are people of the "enthusiast" mining community!)


----------------------------

If we aren't getting very cheap DD-WRT routers in the box to offset the above tethered wattage, Your 60watt number is only for those whom aren't looking at the real numbers.

Without a very cheap low powered controller it is 60 + 20 + 30 = 110 watts. (Low powered laptop)

If a barebones desktop is employed (best case scenario): 60 + 20 + 85 = 165 watts. (not including an active monitor)

I care not to speculate on what a multi-GPU rig consumes.

[Gomeler]

Any chance of decreasing the operating speed and core voltage to boost GH/w efficiency? I fear if BFL delivers on their 1 GH/w efficiency and nobody is close then they will be the only choice for those with a multiple year outlook on using these devices.

[gmaxwell]

If your words are taken literally, then the ASIC sellers are selling people very expensive lemons that will take a year or more to get their investment back....and then.....*barely any profit* per month.

What kind of major business expenses pay themselves back— passively no less!— in just a year?   Perhaps 7%/wk is more to your liking? Those tastes are catered to in other parts of the forum.

There are major risks in mining— including the risk that all that fancy hardware could be worthless before it arrives due to a loss of interest in Bitcoin, security flaws, legal attacks, etc. Or things like breakthrough optimizations for SHA256^2 allowing massive speedups on equal process or someone somehow doing a major run a vastly better process... just to name a few.

Meanwhile y'all have been ordering devices without concrete specs (including BFL customers— a substantial amount of pre-ordering happened when they were saying nothing about power at all!) which makes concrete reasoning about returns impossible if uncertainty about the future difficulty and exchange rate hadn't already made it impossible.

And after that, you're worried about some months difference in payback timeframes for some estimates?  Suicidal.  I hope no one has put money into these asic preorders that they can't afford to lose.

[Tinua]

If a barebones desktop is employed (best case scenario): 60 + 20 + 85 = 165 watts. (not including an active monitor)
--------------------------------------------------------------
Your 60watt number is only for those whom aren't looking at the real numbers.

why you calculate with a single mining device? why not .. with 5 or 10?

60 + 60 + 60 + 60 + 60 + 20 + 85 = 405 watts (not including an active monitor) (for 300Gh/s)
60 + 60 + 60 + 60 + 60 + 60 + 60 + 60 + 60 + 60 + 20 + 85 = 705 watt (not including an active monitor) (for 600Gh/s)

so, what do you mean with your statement above?  

[PuertoLibre]

If your words are taken literally, then the ASIC sellers are selling people very expensive lemons that will take a year or more to get their investment back....and then.....*barely any profit* per month.

What kind of major business expenses pay themselves back— passively no less!— in just a year?   Perhaps 7%/wk is more to your liking? Those tastes are catered to in other parts of the forum.

I can't speak for anyone else, but I am looking for BTC rather than converting those BTC into $$$ directly. So that is not my operating strategy.

There are major risks in mining— including the risk that all that fancy hardware could be worthless before it arrives due to a loss of interest in Bitcoin, security flaws, legal attacks, etc. Or things like breakthrough optimizations for SHA256^2 allowing massive speedups on equal process or someone somehow doing a major run a vastly better process... just to name a few.

Meanwhile y'all have been ordering devices without concrete specs (including BFL customers— a substantial amount of pre-ordering happened when they were saying nothing about power at all!) which makes concrete reasoning about returns impossible if uncertainty about the future difficulty and exchange rate hadn't already made it impossible.

And after that, you're worried about some months difference in payback timeframes for some estimates?  Suicidal.  I hope no one has put money into these asic preorders that they can't afford to lose.

Perhaps there will be a manufacturing defect that will rear it's ugly head after customers have received their orders. Dunno.

Either way, it sounds like one way to stay profitable with mining hardware...is remote upgradability.

If the device cannot be upgraded with better cooling to achieve higher Gh/s then it is a bad option for the long term.

[PuertoLibre]

If a barebones desktop is employed (best case scenario): 60 + 20 + 85 = 165 watts. (not including an active monitor)
--------------------------------------------------------------
Your 60watt number is only for those whom aren't looking at the real numbers.

why you calculate with a single mining device? why not .. with 5 or 10?

60 + 60 + 60 + 60 + 60 + 20 + 85 = 405 watts (not including an active monitor) (for 300Gh/s)
60 + 60 + 60 + 60 + 60 + 60 + 60 + 60 + 60 + 60 + 20 + 85 = 705 watt (not including an active monitor) (for 600Gh/s)

so, what do you mean with your statement above?  

It becomes more efficient with more than one mining device. But...there aren't that many miners out there who are investing 5 thousand or more dollars into their pre-orders.

You could cater to the niche crowd within the niche crowd though...using those numbers though!

[Tinua]

................. But...there aren't that many miners out there who are investing 5 thousand or more dollars into their pre-orders.

Sure???
So, from where does the expected increase of the difficulty about 10x or 20x?

[PuertoLibre]

................. But...there aren't that many miners out there who are investing 5 thousand or more dollars into their pre-orders.

Sure???
So, from where does the expected increase of the difficulty about 10x or 20x?

I am guessing from the collective volume.

Why don't you start a poll thread and ask people what they ordered and the quantity so we have an idea of whom is heavily invested in ASIC mining?

[Tinua]

................. But...there aren't that many miners out there who are investing 5 thousand or more dollars into their pre-orders.

Sure???
So, from where does the expected increase of the difficulty about 10x or 20x?

I am guessing from the collective volume.

Why don't you start a poll thread and ask people what they ordered and the quantity so we have an idea of whom is heavily invested in ASIC mining?

You think one poll is representative? Think not! Furthermore, there is already a list of BFL-order. Perhaps 10% on it

And.......what no one considered so far: The energy prices will rise in the future in many countries, and will not stay on the same level!

[BitSyncom]

The difficulty will be 10x within four weeks of the first ASICs that start shipping, be it BFL or Tom's offering.  After a month, most of the back orders of both companies will be well into consumers hands and the difficulty will rise.  6 months?  It will be a lot more than 10x by then.

How is this any different then? Avalon will be also shipping around that time, and we are bound to finish shipping before you simply due to our much lower volume, are you saying you(BFL) and Tom will ship 150+TH before we ship any of our Avalon units out? then I suppose I do not want to compare my oranges to your imaginary apples.

Either way, I'll be providing a ASIC web based comparison tool within the week and people will decide for themselves.

[Tinua]

Either way, I'll be providing a ASIC web based comparison tool within the week and people will decide for themselves.

Sounds good!

[Inaba]

The difficulty will be 10x within four weeks of the first ASICs that start shipping, be it BFL or Tom's offering.  After a month, most of the back orders of both companies will be well into consumers hands and the difficulty will rise.  6 months?  It will be a lot more than 10x by then.

How is this any different then? Avalon will be also shipping around that time, and we are bound to finish shipping before you simply due to our much lower volume, are you saying you(BFL) and Tom will ship 150+TH before we ship any of our Avalon units out? then I suppose I do not want to compare my oranges to your imaginary apples.

Either way, I'll be providing a ASIC web based comparison tool within the week and people will decide for themselves.

Forgive me if I am mistaken, but aren't you scheduled to ship in January?  Assuming Tom is on time (and there's no reason to think he isn't) between Tom and BFL, yes I expect 150 TH to be shipped by the end of January.  If you're shipping in December, then I retract my statement.  We should have our entire pre-order backlog cleared in January, assuming all goes according to plan. 

As far as the "barely any profit" per month goes, that only applies to devices with GH/w ratios.   The more efficient devices should remain profitable for a very long time.

[BitSyncom]

Forgive me if I am mistaken, but aren't you scheduled to ship in January?  Assuming Tom is on time (and there's no reason to think he isn't) between Tom and BFL, yes I expect 150 TH to be shipped by the end of January.  If you're shipping in December, then I retract my statement.  We should have our entire pre-order backlog cleared in January, assuming all goes according to plan. 

As far as the "barely any profit" per month goes, that only applies to devices with GH/w ratios.   The more efficient devices should remain profitable for a very long time.

Sounds good.

[kaerf]

Assuming bASIC and BFL ship around the start of December.

By the end of Decemeber, I expect bASIC to have shipped all units .

...it will require at least a 7 days of 8-10 hour days,

~800 * 54 = 43 TH/s (assuming bASIC performance stays at 54GH/s..it will most likely be higher)

BFL has more orders and more man power...let's be really conservative and say they ship 75GH/s in Dec.

thats 43 + 75 + 22 = 140 TH/S bare minimum

So that IS ~7X difficulty from day ONE for Avalon customers. I believe your "conservative" ramp up numbers are not conservative enough.

Regardless, I'd say a very "conservative" guess is that within 3 months of Avalon's release, difficulty will be over 10X (IMO, easily). Long terms calculations should begin a minimum of 10X.

Everyone has been saying time-to-ship is the most important factor...well I'd like to look slightly further out than 3 months. The most that I can conservatively plan for is 6-10 months. Who knows what the landscape will be like several months from now.

My outlook for the short-medium term is 10x-30x difficulty. All 3 major offerings will be profitable during this time...how profitable depends on power cost.

Since BFL has published (confident) power numbers...let's use them as a baseline.

Cost per day @ $0.10 @ 60W = $0.14
Cost per day @ $0.10 @ 100W = $0.29
Cost per day @ $0.10 @ 200W = $0.48
Cost per day @ $0.10 @ 400W = $0.96

Cost per day @ $0.25 @ 60W = $0.36
Cost per day @ $0.25 @ 100W = $0.60
Cost per day @ $0.25 @ 200W = $1.20
Cost per day @ $0.25 @ 400W = $2.40

25 BTC/block @ 30M difficulty @ $12 USD/BTC = $12.07 / day

Let's look at how much profit is effected by power

using 60W as baseline
Cost per day @ $0.10 @ 60W = baseline is 0% more profitable
Cost per day @ $0.10 @ 100W = baseline is 1.27% more profitable
Cost per day @ $0.10 @ 200W = baseline is 2.94% more profitable
Cost per day @ $0.10 @ 400W = baseline is 7.38% more profitable

Cost per day @ $0.25 @ 60W = baseline is 0% more profitable
Cost per day @ $0.25 @ 100W = baseline is 2.09% more profitable
Cost per day @ $0.25 @ 200W = baseline is 7.72% more profitable
Cost per day @ $0.25 @ 400W = baseline is 21.10% more profitable

Obviously, as difficulty goes up, profitability goes updown. As such, the affect of power usage on profitable also goes up.

Some people might say 5% is nothing...we tip pool operators that much. Others are trying to maximize every penny. Granted, Avalon now claims a 10% edge in hashing power (over BFL)...I am uncertain how long that edge will last. Both bASIC and BFL have stated that there is (perhaps significant) room for improvement here.


I have orders from each of the big 3, so I'm not trying to bash Avalon here. As a customer I just want a great product that will give me a decent bang for my buck.

BFL offers good hash rate with lower power.
bASIC offers good $/GH...still waiting for power numbers.
Avalon still has good $/GH (with bumped specs)....still waiting for power numbers.

I tried to diversify for the first run of ASICs, but if/when I decide to order more units...I'm looking for the best bang for the buck going forward.

[PuertoLibre]

The difficulty will be 10x within four weeks of the first ASICs that start shipping, be it BFL or Tom's offering.  After a month, most of the back orders of both companies will be well into consumers hands and the difficulty will rise.  6 months?  It will be a lot more than 10x by then.

Congrats, you just made a point that hurts the ASIC sales across the board.

If your words are taken literally, then the ASIC sellers are selling people very expensive lemons that will take a year or more to get their investment back....and then.....*barely any profit* per month.

As far as the "barely any profit" per month goes, that only applies to devices with GH/w ratios.   The more efficient devices should remain profitable for a very long time.

You actually confirmed that....disturbing. Well I applaud your honesty on it at least...

[libertybuck]

> if you can not get access to cheap power, do not go into mining.

严重支持!