Bitcoin Forum

https://forums.butterflylabs.com/announcements/692-bfl-asic-status-2.html#post16043

I figured another delay would more than likely be the case. Maybe they'll start shipping by the end of March if "all goes well"?

I figured another delay would more than likely be the case. Maybe they'll start shipping by the end of March if "all goes well"?

We had expected the bumping to be done by now, as per the previous update(s). That has not been completed yet. There are a number of reasons why this is the case, and we are not pleased with any of them. The bumping facility, which we have no direct contact with, did not complete the NRE on the timeline we had spoke to the packaging facility about.

Week of January 26th

* Leave for fab/bumping facility at the end of the week
* KC facility starts assembly process of units to drop PCB into Units being prepped!

We have changed our bumping facility from the left coast to the right coast (North Carolina), so we may have the 6 wafers overnighted via courier to the bumping facility vs me picking them up and couriering them myself, although I will be meeting them at the bumping facility and making sure they don't sit around for a couple days while some guy eats and egg salad sandwich and wonders why there's a "priority!" box on his desk.

From I will leave for the bumping facility, to I will be meeting with them, to we are not in contact with them (I assume implying third party contractor).

From what Josh has posted, it's the packaging facility which is dealing with the bumping facility.  Changing the bumping facility quite late in the game might have created issues, too.  It's not like all these facilities sit around doing nothing all week just hoping for piddly little rush orders to come in.  BFL might be a large vendor in a very small pond but they're a tiny little fish in the massive customer ocean.

...,although I will be meeting them at the bumping facility and making sure...

I can't tell a customer of mine that their barn wood will be on time even if I have to go to the barn to accelerate the process, when my subcontractor supplier doesn't let me know which barn he's procuring the barn wood from.

I hereby once again challenge Bryan Micon to a 100 BTC or 2000 BTC (or any number in between) wager that BFL will deliver ASICs in 2013 in escrow.

New ETA is end of 2013. LOL.

It sounds like they do know which bumping facility is involved, though, as they made the choice to switch from one in North Carolina to one in California.  While it sounds like there's a hell of a communication breakdown, it also sounds like BFL "didn't know what they didn't know".

There's not really any reason for the packaging facility to assume their client has never heard of a blank alignment wafer and won't be supplying one.  It's possible that the first either the packaging facility or the bumping facility learned about that was when the wafers arrived at the bumping facility.  It's like trying to hurry up the printers when you failed to supply them with the artwork for your job.  Now is the time to make sure that they know exactly what is needed for the remaining steps so they're not delayed by something BFL didn't take into account.

We have changed our bumping facility from the left coast to the right coast (North Carolina),

Are you sure you don't have it backwards?


Unless one's looking at a map of the US from, let's say Oz, the left coast butts up to the Pacific Ocean like California does, and the right coast butts up to the Atlantic Ocean like North Carolina does.

I likely do Phinn, there've been so many changes I've lost track.

Josh has been sucked into giving an estimated shipping date yet again.  Saturday 9 March.

https://forums.butterflylabs.com/bfl-forum-miscellaneous/1163-february-25th-2012-update-discussion-thread-2.html#post16081

It would have been better to refrain from giving an estimated shipping date until testing and packaging are complete.

Now people are going to start asking how many of each unit are going to be in the (reduced) Batch 1, when they're going to get an email about their preferred shipping method, when they should send back their FPGAs, and all that other shit all over again.

If the bumping facility really has the chips bumped and ready for testing by Friday (as they imply for the projected start of deliveries is March 9th) then there is really no need to burn, i.e. destroy 1 whole wafer. 1.000 chips less for the 1st month pre-orders....

I think there point is that there's no benefit in waiting until all of the wafers are bumped - they're going to have to destroy a wafer in order to test unpackaged chips so they might as well destroy the first bumped wafer now as wait until the end of the week.  The alternative is waiting until they have some packaged chips to test.

If the chips test OK then being able to accelerate the delivery of the other 69,000 chips might be worth the sacrifice - but early chip delivery isn't going to worth a damn unless they get the rest of their production processes streamlined in a hurry.  If they can't get 6,000 chips bumped, packaged, placed on boards, assembled into units and shipped, how the hell are they going to manage turning their 63,000 bulk order into working units and getting them to customers before hell freezes over?

..we can send it to the ASIC engineers who can essentially wire bond it manually and test the chips..

Especially not since those 1000 chips could go towards 1st month orders. They are now sacrificing 1st month orders to get later orders filled earlier. As everyone has already paid for their order...and the money is already in the bank for them, they are simply showing a lot of disregard for their early believers.... IMO

How on earth are the engineers going to wire bond a bumped wafer if it's not diced?!

From Josh's latest update:


How on earth are the engineers going to wire bond a bumped wafer if it's not diced?! Maybe Josh should've kept that quarter for himself. You know, to purchase a clue :)

By dremeling the hell out of it.

Expect "We had been promised delivery of the Dremel by today, but unfortunately the Fedex man came and gone without it. I am right now trying to find out what happened to it" real soon now.

Also: "What the hell do you expect me to do? Drive to Lowe's and buy one?"

I know it's not a Dremel, but isn't cutting chips from a wafer a delicate process, one requiring the least amount of heat created?

By dremeling the hell out of it.

Mystery solved! I always thought this was a soldering iron - but it's a dremel  ;D

http://www.bilder-hochladen.net/files/k9l5-s-33e7.jpg

Seems to me everyone here developed stockholm syndrome

Seems to me everyone here developed stockholm syndrome

I thought I was reading about bASIC 2.0 with these new excuses.

think there point is that there's no benefit in waiting until all of the wafers are bumped - they're going to have to destroy a wafer in order to test unpackaged chips so they might as well destroy the first bumped wafer now as wait until the end of the week.  The alternative is waiting until they have some packaged chips to test.

There hasn't been any "drunken" and/or "hacked" BFL posts/accounts yet..

It sounds like they do know which bumping facility is involved, though, as they made the choice to switch from one in North Carolina to one in California.  While it sounds like there's a hell of a communication breakdown, it also sounds like BFL "didn't know what they didn't know".

There's not really any reason for the packaging facility to assume their client has never heard of a blank alignment wafer and won't be supplying one.  It's possible that the first either the packaging facility or the bumping facility learned about that was when the wafers arrived at the bumping facility.  It's like trying to hurry up the printers when you failed to supply them with the artwork for your job.  Now is the time to make sure that they know exactly what is needed for the remaining steps so they're not delayed by something BFL didn't take into account.

The Unknown

As we know,
There are known knowns.
There are things we know we know.
We also know
There are known unknowns.
That is to say
We know there are some things
We do not know.
But there are also unknown unknowns,
The ones we don't know
We don't know.

—Feb. 12, 2002, Department of Defense news briefing Donald Rumsfeld

Has anything ever gone well for BFL? No doubt a lengthy string of really good luck is right around the corner.

Seems strange that they'd delay five months and then choose to render 16% of their first batch of chips unusable to pick up a week or two(hopefully), but they're the self proclaimed experts.

So I really dont see what the big advantage is of destroying 5000+ chips to be able to test, what exactly ?

Goodness no, BFL is terrible at making predictions. I usually take their updates with a grain of salt and add 2-3 weeks onto any date they give. Shipping March 9th? Naw, probably April 6th. ;)

And I agree..killing an entire wafer seems a bit silly, but then again what the hell do I know about chip development. With only 5000 chips, they'll probably only be able to fill June/July orders (assuming they get all 5000).

There's information missing or information being fabricated.

This is BFL we're talking about, so both :D

I sort-of feel bad for Josh. I can understand being in his position; Someone at the mercy of forces beyond his control.

 However, he is quickly becoming the spokesperson for "Bullshit Fabrication Laboratories" and is one bad bender away from pulling a full on Tom and imploding all over the forums.

 Josh/Inaba, as your customer, I advise you to rent a very fast car with no top. And you'll need the cocaine. Tape recorder for special music. Acapulco shirts. Get the hell out of Kansas for at least 48 hours.

 Let this shit with the bump fac resolve itself. Shit happens. But chill the fuck out man.

This is the worst advise ever. Running away from your problems don't make them go away or magically "resolve themselves".

Itchy trigger finger has me wanting to buy 7970's cause the damn price is so high...

What's another month... and another month... and anther month.

*Watches Avalon get shipped*

*watches ASICMiner get built*

Hope to get it by my birthday... maybe Christmas.... 2014. :(

The ASIC team is a third party we have contracted with to do the design and routing on the chip itself, they are located in California.

Do it! Mine, then resell or have a sweet gaming rig!

They are accelerating by 1 week at the most according to their latest timeline.... seeing as the project is already 4 months delayed I really don't see the benefit in destroying 1000 chips

Especially not since those 1000 chips could go towards 1st month orders. They are now sacrificing 1st month orders to get later orders filled earlier. As everyone has already paid for their order...and the money is already in the bank for them, they are simply showing a lot of disregard for their early believers.... IMO

They're bumping orders placed at least EIGHT months ago from the first batch to the second batch with a questionable decision and as we already know BFL can generate delays from nowhere so there's no guarantee that those sacrifices will yield any streamlining going forward at all.

Bullshit Fabrication continues.

"The ASIC team has promised me pictures of the wafer tomorrow, Tuesday the 26th. As soon as I get those, I will be posting them."

26th has come and gone with no update.

Bullshit Fabrication continues.

"The ASIC team has promised me pictures of the wafer tomorrow, Tuesday the 26th. As soon as I get those, I will be posting them."

26th has come and gone with no update.

Josh should probably start saying "mañana" "unemployment sucks" given how often "tomorrow" turns out to be at least a week later than estimated.

It appears that some of the rumors may be true, some requests for refund are indeed being delayed or are ignored for a lengthy period of time.



Source Link: https://forums.butterflylabs.com/pre-sales-questions/1127-cancelling-my-order.html
https://forums.butterflylabs.com/pre-sales-questions/1166-why-hell-your-e-mails-skype-phone-dead-customers-why.html

It appears that some of the rumors may be true, some requests for refund are indeed being delayed or are ignored for a lengthy period of time.

[img]http://i50.tinypic.com/2w1zl6q.png[/ img]

Source Link: https://forums.butterflylabs.com/pre-sales-questions/1127-cancelling-my-order.html
https://forums.butterflylabs.com/pre-sales-questions/1166-why-hell-your-e-mails-skype-phone-dead-customers-why.html

Is it your job to stalk the BFL forums?

I sent their Cust Support an email about an issues (not a refund, however), and it took them 12 days to get back to me. I would assume that a refund would take about the same amount of time, so I don't consider this to be a purposeful delay. You can calm down on the OMG BFL ISN"T REFUNDING ORDERS for a lil bit.

I have not been following this very much, but are they only onto testing chips now?

PuertoLibre stalking the BFL forums.........no way!   LOL

Is it your job to stalk the BFL forums?

I sent their Cust Support an email about an issues (not a refund, however), and it took them 12 days to get back to me. I would assume that a refund would take about the same amount of time, so I don't consider this to be a purposeful delay. You can calm down on the OMG BFL ISN"T REFUNDING ORDERS for a lil bit.

PuertoLibre stalking the BFL forums.........no way!   LOL

What I wonder is if you're not throwing $20-$50K into ASICs why you would be spending so much time posting in threads for an item that you don't even want.  I mean there's educating the public and then there's just making into a life goal.  The same can be said for people trolling in the Avalon threads.  Only reason I come here is to just get updates I missed without having to scan all the BFL forums.

What kind of jobs/lives must some of you have to be able to afford this much time online?

I have not been following this very much, but are they only onto testing chips now?

Composite picture of Test Chips:

http://i48.tinypic.com/35k7mdd.png

 Full Size: http://i48.tinypic.com/142g21g.png

I'll give their "Asic Team" one thing - assuming that really is their die, that's some damn nice wire-bonding.

Enigma

Hmmmm.  That's very open-ended.  Even though they never enforced the old policy, at least it stated that orders were considered non-refundable up until delivery or 1 January, whichever came first.

While not diced, the picture of the full wafer does show the separation of the individual dice.. They look awfully 'square' to me.. This single die looks rectangular.  Optical dillusion, perhaps?

Enigma

Strange indeed. I count about 35 by 35.
Maybe that´s why the image appeared split in the post?
No EXIF data again.

https://forums.butterflylabs.com/dbtgallery.php?do=gallery_image&id=589&gal=gallery&type=full

If you look carefully you'll notice that image is composed from several pictures. That's because they were shot under microscope (maybe bonding machine optics).

How much do they pay you to brown nose for them?

Colour me confused. I am not sure either of the pictures are actually from BFL ASICs. I think the CAD printouts that BFL posted earlier showed their die a little bit but noticeably rectangular. Not as square as the dies on the wafer picture but neither as heavily rectangular as the wirebond picture does. Also the CAD die pictures I think showed the pads for connections to be alongside the edges of the die as I think is the common way of doing things. The wirebond picture shows the bonding pads on the "die" to seem quite strange. I am not sure the wirebonding picture contains any die at all. Maybe it is just some practicing bed for the wiring or something else. If I make a google image search for ASIC wirebonding I get a ton of pictures and none of those I have any doubt at all that they contain a die like the BFL picture I find myself doubting. weird.

Colour me confused. I am not sure either of the pictures are actually from BFL ASICs. I think the CAD printouts that BFL posted earlier showed their die a little bit but noticeably rectangular. Not as square as the dies on the wafer picture but neither as heavily rectangular as the wirebond picture does. Also the CAD die pictures I think showed the pads for connections to be alongside the edges of the die as I think is the common way of doing things. The wirebond picture shows the bonding pads on the "die" to seem quite strange. I am not sure the wirebonding picture contains any die at all. Maybe it is just some practicing bed for the wiring or something else. If I make a google image search for ASIC wirebonding I get a ton of pictures and none of those I have any doubt at all that they contain a die like the BFL picture I find myself doubting. weird.

^^^
This is what I've been trying to find out. Crazyates is such a fanatical BFL apologist and cheerleader that he *has* to be on the payroll.

Probably for the same reason the above people spent months bashing bASIC and Avalon in their own threads.

Pot calling the kettle black as it were.

I've actually stopped even browsing thru most of the Avalon and bASIC threads, let alone taking the time to post in there. Why don't you try it? Just simply ignore every thread here on BTC with the letters BFL in the title. I think everyone will end up a bit happier.

The pads you see are the ones in the green area, which are normally not exposed on a packaged chip. Only on a naked die.

I think that was first design for QFN which they said it failed because of thermal issues

I think Josh said they managed to make the impossible and somehow convert old design to a flip chip (check wikipedia).

This is why I am so confused because what I see there is not a naked die. It looks nothing like a naked die. What is perhaps possible is that there is some layer of stuff on top of the naked die which contains then the pads that are visible on the picture. I understood that this extra layer is created during the bumping process but this wafer is taken aside from the bumping which has not actually started yet and as you can see from the wafer picture it does not contain any of that extra stuff but the wafer picture looks like a naked die. The dies on the wafer picture and the "die" on the wirebonding picture are totally incompatible to each other as their general look. So there is a discrepancy to the situation so that does not make sense from any angle you look at it.

And then there is the additional discrepancy on the square/rectangular issue on top of that.

I dont think you know what a naked die is, or what it looks like. What BFl posted does make sense however, they wirebonded the underside of the die to some test package. Everything you see in the yellow photo is normally covered by black epoxy/plastic/whatever.
In final production chips they will add solder bumps instead of those wires and package it in some substrate which is larger, and can have a (very) different form. Thats when it looks like a "chip".

None of this has nothing to do with the "extra layers". Thats at at the fab, where they halted production to allow for limited changes in case the test chip turns out bad.

They cannot wirebond the underside of the die. That is raw silicon and the uderside contains nothing interesting or useful to anywhere. There is the stuff created in the fabbing of the die and that is all on the topside of the die. In which way the topside of the die is placed to the package (up or down) and what epoxy is put anywhere is irrelevant to this discussion.

Both the wafer picture and the wirebonded picture they should show the topside of the die that is the only side of the die that contains any functional elements of the ASIC. There is nothing on the other side of the die that is interesting in any way.

This is basically what they made to test:
http://www.onsig.com/produc8.gif

This is how it will look like in production:
http://www.emeraldinsight.com/content_images/fig/2190090207005.png

With this new information I am not even sure the CAD files had anything to do with BFL ASIC.

Forget top/down. What I mean is that the feature-side of the die shown on top on the wirebonded picture does not look like normal feature-side of a die like one shown in the wafer picture and that is a problem to explain why is that.

That is correct. BFL originally said that refunds would be allowed "60 days past target", that is 60 days past Oct 31, that is after Dec 31:
https://bitcointalk.org/index.php?topic=87934.msg993703#msg993703  (the only reason I have the link is because I had saved it)

Strangely, the above post was deleted(!)

While not diced, the picture of the full wafer does show the separation of the individual dice.. They look awfully 'square' to me.. This single die looks rectangular.  Optical dillusion, perhaps?

Composite picture of Test Chips

Side by side comparison.

http://i48.tinypic.com/35k7mdd.png http://farm9.staticflickr.com/8236/8517488290_dacd8ffee7_n.jpg

Why was time taken to present the photo on the left when clearly it's a compressed image of the one on the right?

Side by side comparison.

http://i48.tinypic.com/35k7mdd.png http://farm9.staticflickr.com/8236/8517488290_dacd8ffee7_n.jpg

Why was time taken to present the photo on the left when clearly it's a compressed image of the one on the right?

Those pesky 38-pin packages, man.

Now we wait to learn how many of each unit BFL will manufacturer from the first 5,000 chips.  No matter what the answer, it's going to piss off some people everyone.

They should probably hurry up and get their order management software fixed. How are people supposed to update their shipping method when the rate calculator is returning stupid amounts which make no sense?

I also wonder how many of those who sold their pre-orders way back when have moved address by now.  BFL has always said they'll only ship to the address of the person who originally placed the order, so I expect we're going to see a couple of threads from people who bought a pre-order from someone else and never receive their unit.

That 5000 will turn into less than 4000 when the defective chips are excluded. So yeah, after all this time waiting a whole lot of people are just going to be waiting moar and they'll be understandably displeased.

Care to explain why you expect the BFL chips to have a 20% failure rate?

Care to explain why you expect the BFL chips to have a 20% failure rate?

Umm...you want me to explain wafer defects?

Their competition didn't go with 110 and 130nm for their first products because they mistakenly thought higher numbers were better. ;)

Just like the thickness of sheet metal--the higher the gauge, the thinner the metal.

Umm...you want me to explain wafer defects?

Their competition didn't go with 110 and 130nm for their first products because they mistakenly thought higher numbers were better. ;)

Now we wait to learn how many of each unit BFL will manufacture from the first 5,000 chips.  No matter what the answer, it's going to piss off some people.

I understand they expect some chips to fail, but don't you think 20% is a bit high?

BFL has committed to their 1/3, 1/3, 1/3 shipping plan for the first batch.  I'm currently thinking the fairest way to follow this plan would be to split the chips into 3 the three groups and build whatever units are required to fulfill those orders.

28 Feb 2013 Update

The ASIC team has been working all day on testing the various aspects of the chip. There have been no problems so far and all 16 engines hash properly at 250 MHz. Full range tests are scheduled for tomorrow to see how high we can the engines past 500 MHz (the nominal speed the chips are slated to run at).

No update on the bumping at the moment, but I should have one tomorrow as to the disposition of that.

They have 4 units.  It's possible (maybe even likely) that 5,000 chips won't even fill all the orders from the first batch.

The third group isn't really a group as such - they'll just randomly fill orders from either the new order queue or the upgrade queue out of sequence.

The likelihood of someone getting their order filled in the first batch of deliveries still depends on how many of each type of unit BFL decides to make.  The "fairest" way would probably be to devote 1/4 of the chips to each type of unit, but this would mean only 6 mini-rigs shipping in batch 1, which would understandably piss off those mini-rig customers whose Batch 1 orders weren't filled and it's probably preferable to piss off a $1,300 or $150 customer than a $30,000 customer.

Ideally, they should build units so that roughly equal order numbers/order dates are reached by the different product lines.

I was under the impression that their overly complex 1/3 plan was for first month orders only, wasn't the little single introduced much later anyway?

BFL will produce X number of units per type (Jalapeno, LittleSC, SingleSC, Minirig SC) (We do not yet know how many X will be).

We will then ship:

1/3 of X number of units to those that ordered NEW orders.
1/3 of X number of units to those that ordered UPGRADED orders.
1/3 of X number of units will be randomly selected, based on order number, from all the orders received between June 23rd 2012 and July 22nd 2012.

For example, if we have 27 units of Jalapeno's, 18 Units of Single SC and 9 units of MiniRig SC, we would ship:

Jalapeno:

9 units to people who made orders for new equipment, starting with the lowest order number.
9 units to people who made orders for an upgrade for the pack of Jalapeno's coming from a Single.
9 units would be shipped to a randomly selected set of order numbers placed in the first month of pre-orders.

Single SC:

6 units to people who made orders for new equipment, starting with the lowest order number.
6 units to people who made orders for an upgrade from a Single, starting with the lowest order number.
6 units would be shipped to a randomly selected set of order numbers placed in the first month of pre-orders for the Single SC.

Minirig SC:

3 units to people who made orders for new equipment, starting with the lowest order number.
3 units to people who made orders for an upgrade from a Minirig, starting with the lowest order number.
3 units would be shipped to a randomly selected set of order numbers placed in the first month of pre-orders for the Minirig SC.

The numbers above are all hypothetical and we have not determined the actual quantity that will be shipping in the first batch. We are still assessing how many units we are able to assemble per day and how many we will have already assembled and tested prior to our first shipment.

Some people wonder why we are doing it this way. We are shipping it this way, instead of piecemeal because the first people to receive BFL ASIC technology will have a decided advantage for a maximum of 2016 blocks. We want to avoid giving preferential treatment to any given segment of the mining community, providing a level and fair playing field for as many people as possible. We also want to distribute our technology as widely as possible so that no single entity can gain a potentially threatening amount of hashing power that may lead to an abuse of the bitcoin network or blockchain.

** Since the "Little" Single SC unit is a new product, not included in the first preorder phase, all SC orders up to and including orders placed on September 30th, 2012 will be assembled and shipped prior to shipping any NEW orders of the "Little" Single SC. UPGRADE orders will be shipping in the order received for your original order that you upgraded from, 1/2 for FIFO and 1/2 random from all orders.

Yep, but some people upgraded from Jalapenos to Little Singles, so the 1/3 upgrade thing applies because you hold your original order date when you upgrade.  No new Little Single orders will ship in the first batch, though.


Yes, it's overly complex.

That's probably not going to work either, if any significant amount of mini-rig orders were placed very early.  As you said, the little single didn't even exist when pre-orders first opened.  If you go solely by payment dates, it's possible that the whole 5,000 chips would be exhausted before a little single order was reached, so the only hope someone with a little single would have of receiving their unit in Batch 1 would be if their order was selected in the "lottery" where place in the order queue doesn't matter.  By adopting the 1/3 shipping plan, they've committed themselves to building at least some little singles.

I don't think there's any "good" way for BFL to do this.  I think they realise this and it's one reason they've been reluctant to say how they're going to handle this issue.  There's going to be a backlash no matter how they decide to handle it.

I think that FIFO is how they should have handled it - even if it meant that the majority of the first 5000 chips went to mini-rigs - but sticking with the 1/3 shipping plan has needlessly complicated things, especially given that batch 1 is so much smaller than originally envisaged (it was supposed to be 12,000 chips).

Hmmm doesn't seem like they made this date either... are there a lot of these posts where they miss a deadline?

Hmmm doesn't seem like they made this date either... are there a lot of these posts where they miss a deadline?

10 May 2013 Update


Another busy week here at the labs. We continued to refine the LS/S/MR boards while waiting for our next delivery of chips. We expected them today, Friday the 10th, but the packager has been dragging their feet in the process (and one of the reasons we initiated the switch to a new packaging facility, as mentioned previously). The first bulk shipment from the new packaging house is expected on Tuesday and should consist of several hundred chips, which we will be building into all the product lines, assuming no snags are encountered this weekend on the latest rev of the LS/S/MR boards. If there is a blocking problem, we will build them all into 5 GH/s units, but we don't expect any blocking issues at this point.

The rest of the wafers are expected at the bumping facility the following week, and once bumped, they will head to the new packaging facility which is capable of a much higher volume, so we expect them to be packaged pretty quickly and ready for use after they arrive at the facility.

we will see Tuesday .

we will see Tuesday .