Topic: High Efficiency FPGA & ASIC Bitcoin Mining Devices https://BTCFPGA.com - page 39. (Read 218478 times)

Give Inaba a chance to respond,


If he wants to make the bet than we can figure out all the particulars.

Otherwise its a waste of time (like arguing over simulated electrical usage)


and yes in response to crazyates he is allowed one non-vulgar, non-abusive post to confirm or deny his entry into this bet.

I vote that as an exception to point #2, Josh be allowed to post one more time in this thread to confirm his acceptance of this bet, and nothing else.

I would also be wary of input voltages as well. Unless the bASIC is run off of 14V as well.

its very simple

I'm talking once BFL SC devices start to ship - when the masses start plugging them in and mining with them

if they are getting 1.1 Watts per 1Gh/s or less that Inaba wins the bet.

If they are using more than 1.1 watts per 1Gh/s than I win the bet.

It does not get any simpler than that.

Plenty of BFL customers already have kill-a-watts ready to go - it should not be hard to determine what the actual electrical usage of these devices is, once they start shipping.

This got me to thinking, aside from the obviously huge gains to be had from process shrink, do you suppose any of the current ASIC designs use memristor technology?

http://en.wikipedia.org/wiki/Memristor

Maybe my understanding is insufficient, but is it not reasonable to think that more efficient circuits could be created with the availability of a "new" component type?

54Gh/s bASIC Bitcoin Mining Devices
Pre-Order Yours Today!     
Only $1069.99 ! @ http://www.BitcoinASIC.com

@ Cablepair

That is a crazy amount of money...

Therefore you should be extremely specific as to the nature of the electrical use.

When you cite 1.1 watts of electrical use, you should specify WHERE the electrical current will be measured and HOW. I know it sounds stupid of me to bring this up but these are points that the loser and winner would look over the fine print.

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

For example,

Inaba might choose one of his BFL Single SC that has premium chips and actually does 1.1watts per GH/s. So you might want to specify if this is from an average sampling or if Inaba (BFL_Josh) can choose which which sample he wants. (Not all rigs are equal at the chip level.)

-----------

Another point that might later be brought up is WHERE the measurement occurs. At the Chip Socket level?
The Power supply electrical socket? You won't know if the BFL Single SC is using something like a nanoATX power supply with an efficiency of 65% or if it is a quality PSU of 95%.

You should be very specific as to where the measurement occurs and what kind of measurement device is used. (Lab quality or if it is something like a Kill-A-Watt meter.

-----------

You should also specify WHO will do the measurement and how many samples are necessary to verify the electrical draw. (Perhaps you can send each other samples or ask a third party for verification? (Youtube Video recording of the measurement process without editing?)

It might make good PR or bad PR depending on how the details come together.

Oh and last but not least, whether "Tethered wattage" is considered part of the final result or not.

I got my ModMinerQuad in the mail today (once I paid my 10% to the taxman ), and it's hashing away. Great product, I'm hoping the ASICs are just as plug and play.



Great contest, thanks for the preview of my coming bASIC order Tom.

Well this is getting interesting

So this bet would mean no more than 66 watts for an SC single

60 GH at 1.1 GH/W

no more than about 5 watts for a jalapeno.

1650 watts for a mini rig.

ill be away for awhile, let me know if your interested Inaba, and please let's keep this a friendly gentlemen bet. No need for trash talking - a simple yes or no will do.

bbl

*Cough*
http://betsofbitco.in/item?id=782
*Cough*

Seriously though, that's bold, I like it 

And that's the definition of putting your money where your mouth is.

IMO

I think it is completely pointless to argue about simulated (guessed) electrical efficiency.

NOT ONE SINGLE ASIC Manufacturer has offered ANY proof what so ever as to the electrical usage of it's ASIC device.

THAT INCLUDES BASIC, bfl, AVALON, Or any of the others.


So let's put our money where our mouth is shall we?

According to this:

https://forums.butterflylabs.com/showthread.php/16-Announcement-BFL-ASIC-Release-specifications

BFL's line of ASIC devices will consume 1 Watt of Electricity per 1Gh/s of Bitcoin mining speed

(Power figures are +/- 10% of those listed while we finish optimizing our power subsystem.)



Inaba, I will personally bet you 1000 BTC that BFL's line of ASIC devices will consume more electricity than 1.1 watt per 1Gh/s of Bitcoin mining speed.
( .1 or (10%) to compensate for that power subsystem optimization)

I am willing to place the bet immediately and deposit the coins with a reputable person to act as escrow however the terms of my bet are as follows:

1) When BFL's line of ASIC's hit the wild they must not consume more than 1.1 watt of electricity per 1Gh/s of Bitcoin mining speed, and according to everything you are telling us 1.1 watts per 1Gh/s is the most electricity they will use.

2) You will no longer post in my thread, if you post in my thread again you automatically lose the bet.

Let me know if your interested in this bet. if your as sure of BFL's efficiency as you say you are than this should be an easy 12k win for you - like I said I'm ready to deposit the coins in an escrow as soon as you are.

Let me know what you think, even if I lose the 12k it would be worth it - to get your rotten attitude out of my thread for a little while.

Have a nice evening.
Tom

Ok Inaba.

When your fixed cost is extremely low and your initial capital outlay is high, that fixed cost becomes largely irrelevant.

It doesn't matter anyway, because you're deluded if you think any ASIC which will ship in the next couple months will be competitive with future products in three years.

Try doing an Internal Rate of Return and plug in your expected device power usage and then do one with power doubled. You'll see that when all factors are considered, electricity is largely an irrelevant cost.

Wanna bet?

You have got to be kidding me.  You think there's no difference between making $4 per month vs $14 per month?  Really?

Power usage is *everything* when it comes to ASIC.  If you think it's not, you have no grasp on the economics of mining.  ASIC is not CPUs, it's not GPUs and it's not FPGAs.  It's the end of the line for mining technology for the foreseeable future.  What you buy today is what you'll be using in 3 years so long as it's still profitable.  That profitability is determined solely on how power efficient it is. 

If you buy an ASIC device that uses 2x the power as another device for roughly the same price, you might as well throw most of you money into a hole and burn it. 

Belaboring the wattage for ASIC is silly as it will be largely irrelevant as the cost of electricity is practically static for most people.

Let's say an ASIC device costs you $20 a month to run. Well if you're only mining 2 bitcoins per month let's say at today's price($12 or so) you're only making $24 a month - 20 = $4 a month.

So let's say a competitor sells a device that's twice as energy efficient ? woopee you now make $14 a month.

The whole deal with ASIC is the initial sunk capital cost and how quickly difficulty will rise, coupled with the price of bitcoins when/if you sell them.

Electrical cost is largely irrelevant.

The biggest threat to return on investment is difficulty skyrockets and/or bitcoin price crashes.

The other threat is manufacturers getting into a price war and selling 50-60 GH for like $200. This of course ties into difficulty.

A great ASIC vendor would allow previous buyers to have a right of refusal on future batches if there is a significant price drop, or a deeper discount if you are already a customer. This would be the best way to help protect your initial customer's investment.