is butterfly labs mining with those ASICs at the moment?

[Desolator]

I've been watching the charts and the increase in total computation speed seems a bit excessive even at current exchange prices.  Take a look. They're nuts!  It seems to have shot up significantly right around the time that Butterfly Labs would be mass testing some of their 7000+ pre-ordered ASIC chips.  Hmmmmm

They say vague things in their posts on purpose but given their anticipated ship dates, trust me, at this point they're definitely done manufacturing them and are working on testing and packaging them.  You don't finish a 1st release device and cross your fingers and hope it runs once it gets to the customer.  You run it for 24+ hours to make sure defective ones don't go out So I do wonder if they're running many, many GH/s worth of them at once to do a bit of a burn-in or mass reliability test.  And why crunch dummy data when you can make money mining?

If circumstantial evidence isn't enough, try simple business logic.  The price is set at $149 for a 3500MH/s chip. At current prices, they would make $149 after about 10.8 days of mining!  So why doesn't Butterfly Labs just run them for 10.8 days instead of selling them?  Then they'd have $149 and the chips lol.  I mean seriously, read those last 2 sentences again, people.  You think they're not mining with them?  I can't be the only one that realizes this.  From a business perspective, if you had an investment that would pay off 200% in a month and you sell it off instead, you're a crazy person.

So I think a huge portion of them are fully assembled and running right now.  The question is, under these circumstances, when would they stop mining?  Simple. Keep mining until one of 4 things happens:
1. someone notices the potential ability to do a >50% of total computational power attack as that would cause everyone to jump ship and crash the price like the Hindenberg.  By the way, they *potentially* have over 30,000GH/s worth of mining hardware in inventory at the moment and everyone else in the world cumulatively has 21,000.
2. "At current prices, they would make $149 after about 10.8 days of mining" <-- until this is no longer remotely true for any reason
3. someone posts #1 on a bitcoin forum
4. everyone gets super, mega, raging pissed off that the devices are taking too long to ship

So they're either waiting to get closer to the split (see #2 above) or for all their mining to drive the price down slightly in a roundabout way (also #2) or any of the other reasons.  All that pre-order money could easily fund the manufacture and hold of a TON of inventory for a month or two so I seriously think that gigantic spike in the charts is primarily them mining.  They have slightly over $1 million USD worth of preorders right now by the way.  That'd buy the chips, hubs, PCs, internet, electricity, and building for a gigantic mining operation

Of course, they'd never ever admit this.  Bitcoin mining is an arms race.  First it was CPUs then GPUs then FPGAs now ASICs and each were like 10x faster than the last, forcing everyone to buy them.  It's just like guns vs grenades vs nukes.  Since mining is competitive and you only get your share of a certain amount of coins, well back to the analogy in that deals tend to go sour when you find out that the arms dealers are also using the weapons TO SHOOT AT YOU BEFORE YOU BOUGHT THEM!!! ROFL.

So this is all kinda scary, bad news but the good news is, if I'm right that means when the Jalapenos and other ASIC models ship out, the total speed will drop for like a week then go back to a similar level to what it is now

[theasian09]

I had the same thought yesterday and honestly i would not doubt it.  If you need to test the units anyways, why not make some $?

[jgarzik]

Or...

When bitcoin value goes up, the incentive to mine increases.

Just had a big bitcoin conference, with interest spreading.

But most clearly, if BFL's ASICs are anything remotely to spec, mining with even 500 would be far more noticeable than the current hashpower increase.

[chrisrico]

I had the same thought yesterday and honestly i would not doubt it.  If you need to test the units anyways, why not make some $?

I thought I had read that they would take a bitcoin address (or pool credentials) from customers and use that for mining.

[StLucifer]

i would do it too.
this is where ethics will come in to play.
IF BFL truly wants BTC to succeed, they will never run enough units to take over 51% of the hashing power at any one point.
Also, they WOULD take the profits from the mining, no doubt. But considering the magnitude of profits, i believe they would donate a sizable chunk to various bitcoin related projects, ( that is IF they want bitcoin to succeed on a larger scale )

they would not run more than a certain limited number of units at a time.
they would not run them any longer than required to ensure that they are working fine and will not malfunction after shipping.

the question here is, how ethical is BFL?

[conspirosphere.tk]

The logic make sense, but the network hashrate is less than 100% up compared to july (and price fully 100% up).
I expected something quite spectacular from ASICS.

[Stephen Gornick]

I expected something quite spectacular from ASICS.

Exactly.  Until you notice it on the following chart, this additional hashing could just as easily be coming from existing miners that are picking up the GPUs sold by other miners who are liquidating their GPUS to buy FPGAs (or to place orders for ASICs).




 - http://blockchain.info/charts/hash-rate?timespan=all&daysAverageString=7&scale=1

[Grouver (BtcBalance)]

I asked Josh from BFL at the conference.
He said he was sure BFL isn't causing this sudden network capacity peak by testing there ASICs.
Who or what is causing it he doesnt know he said.

[Cablez]

It could just very well be all the mini-rigs that have shipped up til now. There are plenty of them in the wild @ 25GH/s.

[Desolator]

I expected something quite spectacular from ASICS.

Exactly.  Until you notice it on the following chart, this additional hashing could just as easily be coming from existing miners that are picking up the GPUs sold by other miners who are liquidating their GPUS to buy FPGAs (or to place orders for ASICs).

That's what doesn't make sense.  Now is the worst time to pick up GPUs.  Everyone knows about FPGAs and everyone should know about the 50 -> 25 split coming up and should know that ASICs are coming out that will put their GPUs to shame.  So a big July-August rush to add to their rigs seems extremely unlikely.  It's really the worst time to expand a rig and yet there's the graph.

Now I guess more people hearing about BTC who already own a radeon card and simply start mining is very reasonable but still, that's an awfully big increase!

On the other side of things, the jump in August was about 6000GH/s.  That's 1714 Jalapenos, which would be hard to realistically run.  I'd say a company like that could easily come up with 25 testing PCs in 1 room that 1 person could operate alone for testing/mining with sufficient air conditioning, etc.  Prices get nuts with bigger hubs but you can pick up 8-port powered USB hubs and some power strips pretty cheap.  Most boards don't start having issues until you plug in more than 4 USB connections.  Usually cheaper boards have issues using all 14 (H77 chipset has that many I think).  So at 4 each, that'd be 800 units at once.  Keep in mind, they make A LOT bigger devices than the Jalapenos but if they're all Jalapenos, that's 2800 GH/s.  That'd explain half the increase.  Now if you were to take away 2800GH/s from that graph, you get a steady increase at a more similar rate to the many months before it.

Btw, random thing: BFL said they will not sell enough units to 1 individual customer if they suspect they're attempting to >50% attack them.  It's not a 51% attack, as 50.00000001% would do it btw so everyone stop saying that, lol.  Anyway, given the distributed 7000+ pre-orders, they'd need to make about 1.2x more and sell them all to 1 person lol.  That's over 10,000 of their smallest units.  So yeah, no worries there.  The only people that really >50% it is themselves.  But it's stupid for anyone to try that because as soon as you submit bad data to transfer a million BTC to yourself then verify the data yourself with your epic hashing cluster, thousands of people wonder where their BTC just went, everyone panics sells off theirs since the whole system is broken, and they're holding BTC worth $0.01 lol.  Good luck selling it, lol.  Oh, and without the ability to determine ownership, they probably would steal MTGox's BTC bank in the first place which would instantly shut them down.  So double good luck selling their stolen BTC then, lol.  An attack like that would be really, really pointless unless someone simply wanted to shut down the system, not actually make money from the attack.  I have a feeling BFL would be very unhappy if BTC failed so no, I don't think they'll try that

[Desolator]

I asked Josh from BFL at the conference.
He said he was sure BFL isn't causing this sudden network capacity peak by testing there ASICs.
Who or what is causing it he doesnt know he said.

That's what's so insane about that.  They're investing all this money to run a company and design chips and pay their programmers and test everything just to make a profit.  When they'd make $100/week running their wimpiest chip, they're running some of their wimpiest chips.  Just how many PCs and hubs and switches they wanted to throw at it is the limiting factor but I don't think they're running like 4 of them lol.

They just don't want anyone to know that they're competing with all their customers by mining alongside them for big big $$$.

[Grouver (BtcBalance)]

I asked Josh from BFL at the conference.
He said he was sure BFL isn't causing this sudden network capacity peak by testing there ASICs.
Who or what is causing it he doesnt know he said.

That's what's so insane about that.  They're investing all this money to run a company and design chips and pay their programmers and test everything just to make a profit.  When they'd make $100/week running their wimpiest chip, they're running some of their wimpiest chips.  Just how many PCs and hubs and switches they wanted to throw at it is the limiting factor but I don't think they're running like 4 of them lol.

They just don't want anyone to know that they're competing with all their customers by mining alongside them for big big $$$.

I do know Josh also mines for him self on a big scale with the current BFL hardware.
I don't think it's up to me saying how much he has now. I do know though.

[conspirosphere.tk]

That's what doesn't make sense.  Now is the worst time to pick up GPUs.  Everyone knows about FPGAs and everyone should know about the 50 -> 25 split coming up and should know that ASICs are coming out that will put their GPUs to shame.  So a big July-August rush to add to their rigs seems extremely unlikely.  It's really the worst time to expand a rig and yet there's the graph.

Now I guess more people hearing about BTC who already own a radeon card and simply start mining is very reasonable but still, that's an awfully big increase!

I would not say that it doesn't make sense until an ASIC is proven to exist.
Moreover the price increase in the same period probably attracted to mining many with suitable VGAs (gamers) and possibly even investors looking for a quick shot and low risk (since VGA won't lose much value fast whatever happens to BTC), not speaking of motivating marginal miners like me to get back to work, until the going is good.

[Desolator]

I would not say that it doesn't make sense until an ASIC is proven to exist.

They better!  They have $1 mil in pre-orders for them, lol.

I don't think it's up to me saying how much he has now. I do know though.

:-O

[subject]

There's already a post on reddit with this same concern

The top comment by sgtspike brings up a very good point:

"Let's see... would you rather approach a venture capitalist (who knows nothing of Bitcoin), and say, "Hey, I need $2M, and I'll make $12M in sales in a year," or "Hey, I need $2M, and I'll make 1.2 million of this funny internet currency"?
Yeah...
Also, another factor you haven't taken in to account: If BFL was known to have more than half the hashing power of the Bitcoin network (which they undoubtedly would be able to achieve if they produced their own ASIC chips and mined on them before anyone else), all confidence in Bitcoin would immediately falter. And when confidence falters, the price drops, like a rock. And when the price drops like a rock, all those BTC that BFL is mining become worthless. Suddenly, they are out a $2M investment, and it was all for nothing.
They are helping to preserve the Bitcoin network by getting ASIC miners into as many people's hands as possible, increasing confidence, the price per BTC, and, logically, their sales (higher BTC price = more sales)."

[Desolator]

I just said most of that although it reminds me of 1 more reason...maybe they're sick of everyone killing mother earth with millions of watts of GPU power so they they went all earth hippie on us and released something that uses like 100x less power to achieve the same results.  It's possible...are they from California?  

[gbx]

Probably a good way to offset the cost, if they're selling them at a loss...

It's also a good way to burn-in the units and see which ones suffer infant mortality.

[MrTeal]

If this is BFL testing their ASICs, where are they mining?

http://blockorigin.pfoe.be/chart.php

If BFL was adding 6TH/s to the network testing ASICs, you would expect them to be solo mining. If they were running with a pool, EMC would be the natural choice. None of that is seen in the pool block stats; the ratios are pretty close to what they've been before. Slush is down, and 50BTC is 2TH/s up in the last 2 months, but nothing that would conclusively show BFL to be testing ASICs.

[Shadow383]

It could just very well be all the mini-rigs that have shipped up til now. There are plenty of them in the wild @ 25GH/s.

^This. You won the prize!

I can name (just from people who I know have them) ~19 Minirigs, so that's 475Gh/s right there - I doubt BFL manufactured less than 100...

[jjshabadoo]

I also think it's unethical for an employee of a hardware manufacturer to be mining on the same equipment.

Sorry folks, but in a market such as this, it would never be tolerated in other businesses. How do we know Josh doesn't have a competitive advantage ? Is he going to get a 1 TH unit first ?

I think the guy runs a great pool and has been a stand up guy, but once you go to work for a hardware manufacturer, it's kind of bullshit to be competing against your customers in something like mining where timing is everything.

Just my 2 cents. I am NOT saying Josh is doing this and as I understand he ordered his mini rigs before he worked for them.

I just continue to shake my head at how naive BFL seems to be in terms of building reputation. What are they going to sell once the network is saturated and there is no demand ? A more efficient ASIC isn't going to mean shit and if they make a 2 TH rig for the same price then they are screwing their preceding customers. This is really going to be an endless cycle I'm afraid.

On the other hand, building long term relationships built on trust and open information can go along way in a community full of people who are drawn toward decentralization.