1GH/s, 20w, $700 (was $500) — Butterflylabs, is it for real? (Part 2)

[kano]

... and yes my Crystal Ball comment means: no I did not take into consideration any addition to the network or it's effect on the difficulty.

Since that is just as implied by Epoch and others - a guess and nothing more.

also for the remainder of the ~2 week (2016 block) difficulty run it, of course, wouldn't change at all
so if you timed it right with double the month calculation of BFL's (based on the difficulty at the time) you could do it in 2 weeks and not have to worry about the difficulty changing

[1715469522]

1715469522

[1715469522]

1715469522

[RandyFolds]

Assuming nothing else changes; the addition of this many bitforce singles would increase mining difficulty, you would need even more.
It also depends on if your 800MH/s is 800,000,000 or 800*1024*1024.

Here's the number of singles I get after calculating how much higher the difficulty becomes, note this is 800MH; if spec'd at 800MiH/s you need more.

You're confusing mega and mebi, right? 4 294 967 296 "whatevers" is 4.29 G"w", so you're actually calculating MiH/s (you wrote 4 GH/s). But more importantly, wtf? Why oh why would you use this insane prefix? I don't think it has any use outside of bits or bytes, and even in that context I think it's only designed to confuse me...

Exactly I started reading and got utterly confused.  My guess is the math is worthless to.

I think the underlying point is perfect: when you add computing power to the network, difficulty goes up.

Those damn bit prefixes just got me worked up and I had to say something... They're almost as bad as the Imperial unit system.

What have you got against the Imperial system? It makes perfect sense to ride my sixteen-hand horse across a bridge spanning a quarter league over seven fathoms of water. Why wouldn't a single unit be 5280 of the lesser unit? What are you, a commie?

[kronosvl]

... and yes my Crystal Ball comment means: no I did not take into consideration any addition to the network or it's effect on the difficulty.

Since that is just as implied by Epoch and others - a guess and nothing more.

also for the remainder of the ~2 week (2016 block) difficulty run it, of course, wouldn't change at all
so if you timed it right with double the month calculation of BFL's (based on the difficulty at the time) you could do it in 2 weeks and not have to worry about the difficulty changing

actually if you double the hashrate until readjustment those 2016 blocks will come in 1 week (assuming you start with the first block after readjustment)

[RandyFolds]

Due in three days...

Who's got their tracking information?

[deepceleron]

I did actually check to make sure those calculations where correct based on the numbers I used
(Edit: and I chose 800Mh/s as a simple round number since the final figure is still unknown)

The only crystal ball predictions I've been looking at lately are in Zelda Skyward Sword ...

Yeah no idea why he thinks 10^3 isn't 1000
There are no calculations in any of this that would have 1K = 1024
However, of course, 1 difficulty is 2^32 hashes

You see in my first column, all the calculations are in individual hashes, so all is accurate. The second column with Ghash is not used in calculations, I just added that for ease of reading and so you can see what it would translate to.  You see 838,860,800 hash/s is the actual rate for BF singles used in the calculation, I translated the 800 mibihashes/s (2^20). There are many places around Bitcoin where the "computer" MiB, GiB are used - bitcoincharts' network hashrate, for example; lots of stuff is written by coders for coders, so 2^10 is used instead of 10^3, even when it doesn't make much sense. I don't know how miner software is reporting mhash without looking at the code though. Someone selling hardware will probably use the hard drive trick where their 1TB = .909 computer TB.

Is someone going to buy $500,000 worth of these to make BTC valued $1,000,000 in a year anyway...(or somewhere between 1% to +10000% of that depending on how the market goes)

Waiting for first-person reports!

[Epoch]

Is someone going to buy $500,000 worth of these to make BTC valued $1,000,000 in a year anyway...(or somewhere between 1% to +10000% of that depending on how the market goes)

That about sums it up!

[Transisto]

Is there a point to this calculation other than the academic exercise? Using the terms 'bitcoin' and 'per year' in the same context is meaningless. We all know what has/can/will happen to $/BTC and network difficulty in the span of a year.

Forecasting forward even a month is challenging enough even with the benefit of a crystal ball, tea leaves, chicken entrails, and a Ouija board. Let's not forget the halving of the block reward come December ... but that is a separate issue.

If one had that much money to risk, it may make more sense to buy BTC directly and trade it opportunistically throughout the year.

If you're to invest serious money into bitcoin the safest way it though mining.  A BTC at 50$ without a corresponding increase in hashrate make it vulnerable to a 51% attack.

By serious money I mean more than 5m $

/just saying/

[kano]

Is someone going to buy $500,000 worth of these to make BTC valued $1,000,000 in a year anyway...(or somewhere between 1% to +10000% of that depending on how the market goes)

That about sums it up!

The point of my original post was to say that not even BFL would bother to do it (let alone someone else who has to pay more)

It wouldn't be worth the "Crystal Ball" risk - over an entire year who knows what BTC will do and where the difficulty would end up.

Let alone running 710 of them non-stop for a year without any way to get advance warning of failure on them until they simply die and stop ...

Edit: and if anyone wants to get into absolutely correct details ... at average 10mintes a block there's only 4392 blocks in 30.5 days ...
So over a month you'd have difficulty generating more than that unless your hash rate was 10 times the rest of the network.
If it was half of the total network you still couldn't get that many since you'd only get half of the blocks mined - the total blocks would of course be more than that due to the first difficulty change happening faster than 2 weeks for the 2016 blocks but then it would be back to expected difficulty for the rest of your mining and you would only be getting half of the blocks mined per month from then on - or 2196 blocks per 30.5 days

[cuz0882]

"Sales tax will be added if shipped within the United States of America." So what is the American tax rate these clowns are charging anyway?

[simonk83]

"Sales tax will be added if shipped within the United States of America." So what is the American tax rate these clowns are charging anyway?

Try asking them.

[cuz0882]

"Sales tax will be added if shipped within the United States of America." So what is the American tax rate these clowns are charging anyway?

Try asking them.

There is no such thing as United States Sales Tax. It does not exist, kind of like the products they list on the page.

[deepceleron]

What's really worth asking is if they've set up business licenses and reseller license in all 48 states that have sales tax, and registered for a sales tax number, and if it will be properly paid to the recipient's state government, as state sales tax laws require that the buyer's state government gets paid tax. Or if it goes in their pocket...

[dirtycat]

"Sales tax will be added if shipped within the United States of America." So what is the American tax rate these clowns are charging anyway?

Try asking them.

There is no such thing as United States Sales Tax. It does not exist, kind of like the products they list on the page.

now thats funny!

[simonk83]

"Sales tax will be added if shipped within the United States of America." So what is the American tax rate these clowns are charging anyway?

Try asking them.

There is no such thing as United States Sales Tax. It does not exist, kind of like the products they list on the page.

Hey I see what you did there,  that's really clever.

As for the sales tax, I'm not from the US and I understand perfectly well what they mean.   

See what else you can come up with.   Really lacklustre effort so far.

[Costia]

anyone considered that they owe you money for not delivering on time?
its not just that you didnt get the rig itself - they are actually preventing you from mining
If the chips were deliverd on time you would have had already mined quite a lot of bitcoins by now to cover the cost of the systems.
Each month of their delay in delivery of a 50GHash rig costs you 1K BTC = ~5Kusd
Each month of their delay in delivery of a 830MHash rig costs you 17BTC = ~85usd

(the 50Ghash rig looks like something that will burst into flames when you plug it in. Home electric heaters are ~1-2KWatt. this rig is 2.5Kwatt)

[bombo999]

not sure if this had been posted yet

[BTC-engineer]

not sure if this had been posted yet

I saw this picture the first time. Where do you have this picture from?

[heavyb]

asked for a refund from Sonny and got it back the next day. If this is a con, it must be for the long haul.

[deepceleron]

Looking at the circuit board and not investing a whole bunch of time, it looks like this is the likely chipset:

$487
Xilinx
XC6VCX75T-1FFG784C
FPGA Virtex®-6 CXT Family 74496 Cells 40nm (CMOS) Technology 1V 784-Pin FCBGA

Package    784FCBGA
Family Name    Virtex®-6 CXT
Device Logic Units    74496
Number of Registers    93120
Typical Operating Supply Voltage    1 V
Maximum Number of User I/Os    360
RAM Bits    5750784

The spec sheet for current draw also might make one think it would draw around ~22w...

[Matthew N. Wright]

asked for a refund from Sonny and got it back the next day. If this is a con, it must be for the long haul.

It's referred to as a "long con".