Topic: Mini-Rig from Butterflylabs - page 5. (Read 21146 times)

http://www.eia.gov/electricity/monthly/excel/epmxlfile5_6_a.xls

This appears to be net of taxes so rates are slightly understated.  US Residential Jan 2012 average is $.1143

Nice start. Yes, the calculation is highly sensitive to the local price of electricity. Many of the large miners are paying closer to $0.06 to $0.08 per kWh; $0.12 is uncommonly high for North America. The current BFL Single leadtime is 12 weeks (no January order has yet shipped) so July 1 would be a more realistic 'start mining with a Single ordered today' date. Under these conditions the GPU has an even larger lead for break-even.

The other aspect is that the salvage/resale value of those 2 items. The 5970 should retain more of its original purchase price than the Single.

Back on topic ... @BFL-Engineer: if you are ready, can you post some official information on the Mini-Rig on your website? Forum speculation is nice but pointless after some time; if this is a real product offering we'd appreciate some official BFL information/announcement.

OK - I'll bite...

5970 - $400 (trying to skew to BFL's favor with a price on the HIGH side)
 - Conservatively running 700 Mh/s
 - .43 BTC/day
 - $2.16 (using $5 FX as a round number)
 - Could be mining by April 2, 2012
 - 185 day (26 weeks) payback (keeping it simple and ignoring power costs)

BFL Single
 - 832 Mh/s
 - .51 BTC/day
 - $2.57 (again, using $5)
 - Delivery in 6 weeks (BEYOND optimistic) - Start Mining May 14, 2012
 - 233 day (33 week) payback

So by early October the 5970 is paid off and the single, with the 6 week additional leadtime (realistically like 12), wouldn't be paid off in 2012...

If you net out power costs assuming $.12 Kwh.  5970 at 250 W (memory downclocked) and BFL Single at 80 W.

For the 5970, $2.16 a day is reduced by $.72 to net out $1.44 profit. 277 day payback (40 weeks) - so paid off by Week 2 of 2013
For the BFL single, $2.57 a day is reduced by $.23 to net out $2.34 profit.  256 day payback (36 weeks) - so paid off in week 4 in 2013.

So maybe nedbert9 is more correct than originally assumed.  Especially if power costs are high.

But there are three points which skew this closer than it should be...

1.  $400 purchase price of 5970
2.  Ignored the rapacious shipping on the BFL
3.  Took the BFL leadtime to 6 weeks

Need to check the math on that.

Thermal envelope has been reached and I do not have the time or energy to deal with offsite locations...

@Epoch - it is a very fair point you make however...

EDIT - I ahve increased my GPU operations 200% in the last 6 weeks 

Who says I haven't boosted my GPU holdings?  I have, actually... though I am running into space and power issues at this point... and more specifically, coming into summer, GPUs are less attractive since their cost to operate is going to increase dramatically with cooling requirements.  Over the winter, they helped quite a bit by heating the house - I have limited space in my DC, so only a portion of my rigs are there, the rest are in the basement.

Additionally, the absolutely nightmare/headache that keeping a large GPU farm running is also a huge turn off.  I am already required to put more time than I want into maintaining, RMAing, replacing, etc... parts to the GPU farm I have and increasing my GPU holdings would only increase the hassle there.  My sanity is worth a month or two waiting for some GPUs that won't (hopefully) give me any hassle vs trying to wring ever last dime out of GPUs before the block reward halving.  

I think the answer to this relies on one answer: How much of the Bitcoins traded daily in exchanges are fresh mined coins?
Anybody has any figures regarding this subject?


Regards,

If 'time is of the essence' as you seem to suggest, you would be better off going down the street and buying up a whackload of GPUs (setting up a few mining rigs takes a week) rather than waiting 2, 3, 4 months for a BFL product. That 2,3,4 months of time can be earning you BTC, and the GPUs will break even sooner than a delayed Single or Rig/Mini if you start the clock now. Different story long-term, but if you are concerned with maximizing your profit before December then GPUs may be the best way of achieving that.

I have a different theory: when the reward is halved, I agree that it will have little impact on the BTC exchange rate. Most of that is buoyed by speculation, and speculators don't care (and/or are oblivious) about block reward or mining profitability. But miners are not; marginal miners (stealing DeathAndTaxes's terminology here) will become non-profitable and will stop mining. What we will see is not a BTC price increase, but rather a significant reduction in difficulty over the 2 or 3 months following the reward drop. Not a 50% reduction, but significant nonetheless. Marginal miners (those with inefficient GPU setups and/or high electricity costs) will be forced out, leaving the high-efficiency GPU miners and the FPGA crowd.

I tend to agree with Inaba- hence "time is of the essence" and would love to see this be more real world with penalty clauses etc. for >12 week delivery (stick).  Or on the other side, a bonus for each week less than 12 week delivery (carrot).

And yes, you can game it, but you get the gist of it...

I very seriously doubt we are going to see the price go up much when the block reward halves.  There will be a slight increase leading up to it, then some wobbling after the change, then it will settle down again to whatever it was prior to the build up.  The "supply" isn't going to change the demand, since the demand is 99% speculation anyway.

So yes, you need to get to mining as much as you can and get back your ROI/break even prior to the block  reward halving, or you'll be doubling your ROI time when that happens.

Regarding the "EU Distributor", if the problem is the VAT, then opening an EU distributor
won't help since the VAT will definitely exist in that case. There may be a way or two to
get these units via post without incurring additional costs, however, with a distributor,
there won't be any.

Furthermore, an EU Distributor must pay social security charges, income taxes, professional
taxes, VAT (paid by the customer), land tax, business tax, probably CO2 tax... There is a whole
tax theme park behind it...


Good luck,

Called and emailed. He is a very nice guy.

gigavps,  did you actually call Sonny at the tail end of last year for info?  You seem to have been 'in the know' for quite a while. 

As for me, as soon as Inaba posted his stats that we could follow, I put in a order that night!!

I wouldn't call it "ordered blindly". Anyone could have picked up the phone and called them, ask for references, visited them or any number of other things to do some due diligence before making their purchase. I recommended this to multiple people in the old threads regarding BFL.

While I definitely understand that the urgency is different, the products are shipping, which is definitely good.

One thing I understand completely, while not understanding at all, is why people are looking at the block reward halving as why you need to mine now. While yes, mining now gets more reward, mining later will get rewards, and the value of the coins will increase. It's a 2 edge sword, if everyone started getting their BFL orders in 3-4 days, you can sure bet the block halving would happen sooner, the first 10 rig boxes being delivered will immediately add 500GH, and that is very significant.

The most I can say is those who gambled with BFL and ordered blindly, won that gamble. Right now, BFL is the option for best price/performance in the FPGA world. I certainly hope to see some ASICs soon!

OK I'll risk joining this discussion before it turns into a brawl... I was one of the northern Europe customers who ordered a Wicked Lasers Arctic when they announced them. Now Wicked Lasers are rather good at marketing, and the hype was great. The emails reached the right people, the right forums went viral, etc. and WL stood looking at a monstrous number of orders - a hell of a lot of them from the USA.

Of course, the original product was, IIRC, illegal in the USA for virtually any contrived reason. It looked like a light sabre toy but put out 1W of laser light - it was weapons grade and shipping with a nice set of goggles wasn't going to cut it. So a complete design respin was required to follow US weapons regs (which, for lasers, were being thought up on the fly). On top of that, the original production process (rip the diodes out of projectors and rebuild into fancy cases) wasn't scalable to the necessary level and the source of diodes wasn't enough.

Product didn't ship, customers got angry, WL themselves lost control of their own order management system (I received three Arctics for some reason - and yes, I offered to pay in full for all unexpected, unordered product). But product eventually shipped. I still have an Arctic and it works well. Product quality didn't disintegrate (at least in my tiny sample).

The difference with BFL is that time risk isn't a matter of 'I WANT MY PONY!' - this isn't a puerile desire to have the latest cool kit before everyone else has. If BFL were the only players in the market for orders-of-magnitude-superior-to-GPUs mining kit, then everyone would be in the same boat (apart from those leet enough to design their own bitstreams and commission their own FPGA big-boxes). But they're not. And there's a time limit before the block reward halves, and we don't have an historical analogue of *this* event (re: money supply singularity) so we can't easily predict how the BTC economy will respond to it, whether by raising BTC/xxx FX rates or wiping out mining capacity.

So most miners are looking to ensure their investments pay off *before* this event. And I'll admit I've got a dog in this fight, as I'm an FPGA miner too, but not in the USA so not a BFL customer. So call me biased, I don't mind. I like their 'plug and play' approach, but you can be sure that I'm hoping they take their time as *long* as possible to 'get the product right' After all, if anyone with a few tens of thousands of dollars and a decent electrical supply can buy a box, plug it in and become a 'player' in the mining world, it removes the current technical barrier to entry, which is the only advantage people in high-power-cost locations (like both of mine) have right now. Mass mining centralisation to a few players with free power and unlimited capital is *not* good for the Bitcoin economy.


(and I would normally say, 'BFL - get an EU distributor for your products FFS', and then be polite and say 'please'... but it very much sounds like they're far too overstretched as it is to accept MORE customers and orders. This is a USA thing, and I desperately hope they don't revisit the Wicked Lasers scenario and get quagmired by US customs / export regs - after all, IIRC, the US is touchy about cryptographic equipment being exported without limitations and full disclosure... and will BFL tell the government *exactly* how it works if they're not telling us?)

Those poor mining hardwares giving up their life and soul for our greedy gain 

Yeah on second thought I agree with Giga & Epoch.  It seemed "cool" (plug in ethernet and power and it just starts hashing) but having an independent host means if host dies you don't need to RMA you $15K investment.  You also have more control over the OS, monitoring, miner software, etc.

Still I would imagine on a long enough timeline hashing will become essentially an commodity appliance that you never open, upgrade, or modify.   In time it will just be a rackmount box which you connect power and ethernet and then login to a webpage to set a couple configs and it just hashes away until it dies or becomes uneconomical.

It'd be great to have an option of getting a box with those features.

I also do not see a need; an internal host is unnecessary and even undesirable. An external PC of your own choosing, running software of your own choice, controlling mining hardware you have chosen to buy, is preferable. There is value in 'choice', which is something most of us here understand well.

Also you do not need a 100W PC to control these miners (although if you have one kicking around anyway, might as well use it) ... a small, cheap, 20W Atom/Brazos nettop like an Eee Box EB1012/EB1021 is more than adequate.