Butterfly Labs shipping 300 units a day

[geofflosophy]

His point is that they could have made more money, both by charging more per GH and by having more customers as a result of a slower increase in difficulty scaring less people off. I'm not sure I totally agree with this point, would it even occur to many of the people who ended up buying 60 GH rigs to buy 10 smaller ones had they not had the option? With that said, there is clearly a good business in selling smaller units to users while they are profitable then selling progressively larger units to the same customers as difficulty climbs.

Avalon made a huge mistake launching a 60GH/s rig straight up, too much of a leap from FPGA.

From a business point of view, Avalon selling $10 million + worth of chips isn't a bad thing. All they have to do is package those suckers up and ship em. Easy profit.

[1715322930]

1715322930

[1715322930]

1715322930

[1715322930]

1715322930

[1715322930]

1715322930

[1715322930]

1715322930

[1715322930]

1715322930

[k9quaint]

His point is that they could have made more money, both by charging more per GH and by having more customers as a result of a slower increase in difficulty scaring less people off. I'm not sure I totally agree with this point, would it even occur to many of the people who ended up buying 60 GH rigs to buy 10 smaller ones had they not had the option? With that said, there is clearly a good business in selling smaller units to users while they are profitable then selling progressively larger units to the same customers as difficulty climbs.

Granted, their pricing has not been optimal. But, the rise in difficulty is not under Avalon's control. Avalon first priced their units as if BFL was going to deliver in "Two Weeks(TM)" and that turned out to be too cheap given BFL's delays. They then priced their 3rd batch a lot higher and now people are asking for refunds because they fear it won't mine enough BTC to recoup it's cost. This fear is driven by Avalon's delays, BFL & ASICMiner delivering, and Bitfury & KNCMiner making good progress.

Avalon needed to sell their first batch of chips, and that batch had a minimum size. I don't think it really matters if they put 20 chips in a box or 200 chips in a box other than it is easier to ship 300 boxes than it is to ship 3000. Given they were trying to sell a fixed number of chips, the demand exceeded the supply so people would buy 1 66GH/s or 11 6GH/s devices.

Also, it has since become clear that Avalon's core competency is in designing and producing chips, not devices.

[YipYip]

What is really annoying for people like me is that I ordered 3 Jalepenos in the first couple of weeks.  They should have arrived and paid for themselves by now, but I UPGRADED my order to the better mini single, which no one has yet seen!!.

I wish I could down grade my order again.  Id be much better off.

If everyone had just ordered Jalepenos the entire BTC net would have been better off, greed for bigger and bigger rigs will kill the ROI for everyone. Avalon made a huge mistake launching a 60GH/s rig straight up, too much of a leap from FPGA. It started a trend which is already causing problems. They should have simply sold 10 x the number of 6GH/s rigs. for a much better balanced net, and they could have charged more per GH/s.

Many humans suffer from greed, ordering big rigs straight up is an obvious example of this, like dogs competing for a food bowl.

Everything you say is a fucking lie !!!

[erk]

His point is that they could have made more money, both by charging more per GH and by having more customers as a result of a slower increase in difficulty scaring less people off. I'm not sure I totally agree with this point, would it even occur to many of the people who ended up buying 60 GH rigs to buy 10 smaller ones had they not had the option? With that said, there is clearly a good business in selling smaller units to users while they are profitable then selling progressively larger units to the same customers as difficulty climbs.

Avalon made a huge mistake launching a 60GH/s rig straight up, too much of a leap from FPGA.

From a business point of view, Avalon selling $10 million + worth of chips isn't a bad thing. All they have to do is package those suckers up and ship em. Easy profit.

They wouldn't have needed to buy 10 smaller Avalon could have just released TH/s at a slower daily rate and charged way more for it, so ther would have been less TH/ s competition for the miners. Everyone wins.

[glendall]

It's not so much 'fool me once, shame on you, fool me twice, shame on me'.

This is more apt : 'fool me once, shame on you, fool me 308 times, I'm a moron for taking anything this company says as remotely possibly accurate.'

[YipYip]

His point is that they could have made more money, both by charging more per GH and by having more customers as a result of a slower increase in difficulty scaring less people off. I'm not sure I totally agree with this point, would it even occur to many of the people who ended up buying 60 GH rigs to buy 10 smaller ones had they not had the option? With that said, there is clearly a good business in selling smaller units to users while they are profitable then selling progressively larger units to the same customers as difficulty climbs.

Avalon made a huge mistake launching a 60GH/s rig straight up, too much of a leap from FPGA.

From a business point of view, Avalon selling $10 million + worth of chips isn't a bad thing. All they have to do is package those suckers up and ship em. Easy profit.

They wouldn't have needed to buy 10 smaller Avalon could have just released TH/s at a slower daily rate and charged way more for it, so ther would have been less TH/ s competition for the miners. Everyone wins.

blah blah blah ...sockpuppet sockpuppet sockpuppet....


You sound like the IRAQ Information minister....

"invasion...there is no invasion"

"deliverys ...there are hundreds and thousands and all of these people do not deserve to have our products in the first place as they are greedy dogs !" 

You are a fucking joke !

[k9quaint]

His point is that they could have made more money, both by charging more per GH and by having more customers as a result of a slower increase in difficulty scaring less people off. I'm not sure I totally agree with this point, would it even occur to many of the people who ended up buying 60 GH rigs to buy 10 smaller ones had they not had the option? With that said, there is clearly a good business in selling smaller units to users while they are profitable then selling progressively larger units to the same customers as difficulty climbs.

Avalon made a huge mistake launching a 60GH/s rig straight up, too much of a leap from FPGA.

From a business point of view, Avalon selling $10 million + worth of chips isn't a bad thing. All they have to do is package those suckers up and ship em. Easy profit.

They wouldn't have needed to buy 10 smaller Avalon could have just released TH/s at a slower daily rate and charged way more for it, so ther would have been less TH/ s competition for the miners. Everyone wins.

Except the people who get their units later rather than sooner. Those people do not win, they pay the same price for hashing at higher difficulty.

You were not around back in January 2013, but Avalon was racing against a phantom horse named "BFL in Two Weeks". It wasn't until batch 3 that Avalon realized that BFL was not going to ship in two weeks, and then they raised the price of their units.

[DeathAndTaxes]

What is really annoying for people like me is that I ordered 3 Jalepenos in the first couple of weeks.  They should have arrived and paid for themselves by now, but I UPGRADED my order to the better mini single, which no one has yet seen!!.

I wish I could down grade my order again.  Id be much better off.

If everyone had just ordered Jalepenos the entire BTC net would have been better off, greed for bigger and bigger rigs will kill the ROI for everyone. Avalon made a huge mistake launching a 60GH/s rig straight up, too much of a leap from FPGA. It started a trend which is already causing problems. They should have simply sold 10 x the number of 6GH/s rigs. for a much better balanced net, and they could have charged more per GH/s.

Many humans suffer from greed, ordering big rigs straight up is an obvious example of this, like dogs competing for a food bowl.

Nonsense.  If 6GH/s rig would be profitable then buying 10x 6GH/s rig would be equally profitable or renting a small amount of commercial space and buying 1000x 6GH/s rigs.  CPU and GPU rigs are constrained by the cludginess of trying to scale out.  Rigs which communicate with a host over USB lack any such constraint.

Either way the network is going to grow to where it is barely break even for new hardware (including electrical cost) from there it is only going to grow if/when either the exchange rate rises or more efficient devices are produced.

[erk]

Nonsense.  If 6GH/s rig would be profitable then buying 10x 6GH/s rig would be equally profitable or renting a small amount of commercial space and buying 1000x 6GH/s rigs.  CPU and GPU rigs are constrained by the cludginess of trying to scale out.  Rigs which communicate with a host over USB lack any such constraint.

Either way the network is going to grow to where it is barely break even for new hardware (including electrical cost) from there it is only going to grow if/when either the exchange rate rises or more efficient devices are produced.

If Avalon charged $1,000 for a 6GH/s unit, you wouldn't see 1,000 of them going out in one hit would you.

You seem to be missing the point. In order to keep the BTC network balanced and growing at a sustainable rate, you don't release hardware that is orders of magnitude higher hash rate than what's currently available.

There is nothing stopping BFL or Avalon building a 10TH/s unit right now that requires massive 3 phase power, but there is no need.

[k9quaint]

Nonsense.  If 6GH/s rig would be profitable then buying 10x 6GH/s rig would be equally profitable or renting a small amount of commercial space and buying 1000x 6GH/s rigs.  CPU and GPU rigs are constrained by the cludginess of trying to scale out.  Rigs which communicate with a host over USB lack any such constraint.

Either way the network is going to grow to where it is barely break even for new hardware (including electrical cost) from there it is only going to grow if/when either the exchange rate rises or more efficient devices are produced.

If Avalon charged $1,000 for a 6GH/s unit, you wouldn't see 1,000 of them going out in one hit would you.

You seem to be missing the point. In order to keep the BTC network balanced and growing at a sustainable rate, you don't release hardware that is orders of magnitude higher hash rate than what's currently available.

There is nothing stopping BFL or Avalon building a 10TH/s unit right now that requires massive 3 phase power, but there is no need.

You were not around back in January 2013, but Avalon was racing against a phantom horse named "BFL in Two Weeks". It wasn't until batch 3 that Avalon realized that BFL was not going to ship in two weeks, and then they raised the price of their units.

[RoadStress]

Nonsense.  If 6GH/s rig would be profitable then buying 10x 6GH/s rig would be equally profitable or renting a small amount of commercial space and buying 1000x 6GH/s rigs.  CPU and GPU rigs are constrained by the cludginess of trying to scale out.  Rigs which communicate with a host over USB lack any such constraint.

Either way the network is going to grow to where it is barely break even for new hardware (including electrical cost) from there it is only going to grow if/when either the exchange rate rises or more efficient devices are produced.

If Avalon charged $1,000 for a 6GH/s unit, you wouldn't see 1,000 of them going out in one hit would you.

You seem to be missing the point. In order to keep the BTC network balanced and growing at a sustainable rate, you don't release hardware that is orders of magnitude higher hash rate than what's currently available.

There is nothing stopping BFL or Avalon building a 10TH/s unit right now that requires massive 3 phase power, but there is no need.

It's a free market. If Avalon charged $1,000 for a 6GH/s unit what's stopping another company to offer a better deal? (and so on)

[dwdoc]

What is really annoying for people like me is that I ordered 3 Jalepenos in the first couple of weeks.  They should have arrived and paid for themselves by now, but I UPGRADED my order to the better mini single, which no one has yet seen!!.

I wish I could down grade my order again.  Id be much better off.

If everyone had just ordered Jalepenos the entire BTC net would have been better off, greed for bigger and bigger rigs will kill the ROI for everyone. Avalon made a huge mistake launching a 60GH/s rig straight up, too much of a leap from FPGA. It started a trend which is already causing problems. They should have simply sold 10 x the number of 6GH/s rigs. for a much better balanced net, and they could have charged more per GH/s.

Many humans suffer from greed, ordering big rigs straight up is an obvious example of this, like dogs competing for a food bowl.

So people buying FPGAs were not greedy compared to people with GPUs, and people with Jalapenos were not greedy compared to people with FPGAs, but people buying Avalons and Singles are greedy compared to people with Jalapenos?? Hmm let me guess- you bought a Jalapeno...

[joeventura]

Yes 300 units a day indeed.

Here is today's shipping update, copied from BFL at 7pm on 8/17



Most Jalapenos through Oct. 27 have been shipped.

No Little Singles

No Singles

No MiniRigs

[dominicus]

Maybe they're counting each ASIC chip as a "unit", or maybe each pound of shipping weight counts as a "unit".   

Even so, 300 a day sounds high.

It's tough to estimate how many orders BFL is shipping per day.  That being said, just as vocal as forum in bitcointalk is, so is the forum at BFL, sometimes worse.

BFL is reporting the pay-dates they've already shipped by models several times/week.  Today they reported shipping 10-calendar-days worth of Jalapeno orders in a single shipping day.  Not one comment from any customer alleging they were passed up or that BFL was misrepresenting (except, of course, the ones that didn't order Jalapenos, who are now reconsidering their model choice).
https://forums.butterflylabs.com/blogs/bfl_jody/242-wednesday-july-17-2013-shipping-update.html

Should they continue to ramp up delivery speed in this fashion, quite a few folks are gonna be angry they didn't take the risk with a few orders.
I mean, I'm not a BFL lover or hater.  I have money trapped for the better part of this year.  But had I ordered when the Jalapenos were $120 each, and later asked for a refund out of "spite" for their sloooow ship start, I would be kicking myself now.

You're not supposed to be buying ASIC's with money you're going to crawl up the walls later for.

[Phinnaeus Gage]

I don’t even ask what its for, expecting that information on a machine like that wouldn’t be shared with this writer. I was also afraid that my mind might be bitcoin blown.

I guess the author didn't see the video where Josh was trippin' over himself while trying to explain the blockchain.

Zerlan is flying a small remote controlled helicopter around his office, shooting projectiles from the airborne machine into a wall.

It's a damn  shame I had to weed through the entire article to finally come across some substance.

Too bad the followin' image of Zefram flyin' the Huey wasn't included in the article.

[dwdoc]

The fact that they are shipping out sample chips to bulk chip purchasers in a timely fashion suggests that delays are due to issues with other parts and/or assembly and not due to delays in chip acquisition. Not that it's fair but I think BFL chip purchasers will have working clones before many of those who bought miners...

[k9quaint]

BFL is reporting the pay-dates they've already shipped by models several times/week.  Today they reported shipping 10-calendar-days worth of Jalapeno orders in a single shipping day.  

The problem with reporting shipping by "day's worth of orders" is we don't know how many units were paid for on that day. Could be 2. Could be 20. Could be 200. It gives us very little insight as to the rate at which BFL is able to ship units out. For whatever reason, BFL feels it should not reveal how many actual units are leaving the building each day. They claim they can do 300 units a day, but can't seem to make any progress past the first day of orders after 10 days of working on singles and mini-rigs. So did they build and ship 3000 singles in 10 days? Did BFL just add 150 TH/s to the hash rate in the last 10 days? Who knows?

The fact that they are shipping out sample chips to bulk chip purchasers in a timely fashion suggests that delays are due to issues with other parts and/or assembly and not due to delays in chip acquisition. Not that it's fair but I think BFL chip purchasers will have working clones before many of those who bought miners...

Shipping a couple of sample chips is very different than shipping a couple of thousand chips. It is very possible that BFL is running low on chips and the bulk of their customers (as measured by hash rate) will have to wait until they receive the next order (100 days lead time).

[Phinnaeus Gage]

300 Units a day.
9,000 Units a month.
60,000 Units ordered.
~7 Months from July 1 before all receive their units.

They share the building with a general contractor company and are currently in talks about possibly taking over and occupying the entire building. It doesn’t take much to realize that this company is outgrowing its office.

If they did have the whole building, what exactly are they plannin' on buildin' as their third hoorah?

Do you really think Huber is plannin' to vacate his pride-and-joy facility. I'm sure this revelation will come to a shock to the current employees of A.L. Huber.

[erk]

So people buying FPGAs were not greedy compared to people with GPUs, and people with Jalapenos were not greedy compared to people with FPGAs, but people buying Avalons and Singles are greedy compared to people with Jalapenos?? Hmm let me guess- you bought a Jalapeno...

Are you trying to make a point with the string of questions, or just having a pointless rant?  
The hash rate increase from GPU to FPGA was modest 600MH/s ->800MH/s, what it enabled you to do was use several FPGA chips for similar running cost due to the improved power consumption. The jump from FPGA to Jalapeno was significant 800MH/s->5,000MH/s, but not the order magnitude like the jump from FPGA to Avalon batch #1 was, 800MH/s -> 60,000MH/s

[dominicus]

BFL is reporting the pay-dates they've already shipped by models several times/week.  Today they reported shipping 10-calendar-days worth of Jalapeno orders in a single shipping day.  

The problem with reporting shipping by "day's worth of orders" is we don't know how many units were paid for on that day. Could be 2. Could be 20. Could be 200. It gives us very little insight as to the rate at which BFL is able to ship units out. For whatever reason, BFL feels it should not reveal how many actual units are leaving the building each day. They claim they can do 300 units a day, but can't seem to make any progress past the first day of orders after 10 days of working on singles and mini-rigs. So did they build and ship 3000 singles in 10 days? Did BFL just add 150 TH/s to the hash rate in the last 10 days? Who knows?

The fact that they are shipping out sample chips to bulk chip purchasers in a timely fashion suggests that delays are due to issues with other parts and/or assembly and not due to delays in chip acquisition. Not that it's fair but I think BFL chip purchasers will have working clones before many of those who bought miners...

Shipping a couple of sample chips is very different than shipping a couple of thousand chips. It is very possible that BFL is running low on chips and the bulk of their customers (as measured by hash rate) will have to wait until they receive the next order (100 days lead time).

Well, the evidence is they are accelerating the amount of "days-worth" each shipping day, but there's still limited info available.  Is it because order volumes dropped, or increase production rate, or both?  BFL will not release data on amount of orders for each calendar day.  Not even public companies do that, so why crow about it?  To get some indication about the specific ordering period sales rate, one could look at the data BFL customers volunteered at http://bfl.ptz.ro/ :
Period 10/7/12 to 10/16/12: 52 Jalapeno orders registered
Period 10/17-10/27: 29 Jalapenos orders registered
Period 10/28-11/6: 13 Jally's orders registered

Whatever math you do will have tons of fuzziness.  More datapoints of shipping reports may be needed, but signs point to BFL accelerating their catch-up efforts on Jally sales.

Now for a good question:  given the evidence above, if you had 3 BTC to spend, would you place a new order for a BFL Jalapeño able to run at 7GH/s and 65W, or order 3 Asicminer Erupters for immediate delivery running at 1GH/s @ 12W?