Newest Update From BFL on Chips

[dresdenreader]

https://i.imgur.com/aMXYg8l.png

transcript: 08:20 PM BFL_Josh: I got more info today. The bumping should be done by tomorrow and the ASIC team should have sample chips by Friday, or at least that's when they are expecting them (about 40) so they can run tests and make sure everything is good. Packaging should be done around the same time and shipped off to Chicago after that.

[1715225992]

1715225992

[1715225992]

1715225992

[1715225992]

1715225992

[smoothie]

https://i.imgur.com/aMXYg8l.png

transcript: 08:20 PM BFL_Josh: I got more info today. The bumping should be done by tomorrow and the ASIC team should have sample chips by Friday, or at least that's when they are expecting them (about 40) so they can run tests and make sure everything is good. Packaging should be done around the same time and shipped off to Chicago after that.

So does this mean BFL has not done any real tests on their ASIC chips?

And you all thought they were going to ship soon after getting their chips...

Pray to the ASIC Gods that the chips are 100% perfectly working.

Big surprise  

[dresdenreader]

Hey now, don't shoot the messenger = )

I guess we shall see what happens, I'm glad I was lucky enough to grab that screen of him giving us an update.

Also, this happened too - http://mineforeman.com/2013/02/20/kanoi-from-the-cgminer-crew-posts-photos-of-his-butterfly-labs-visit/

[crazyates]

https://i.imgur.com/aMXYg8l.png

transcript: 08:20 PM BFL_Josh: I got more info today. The bumping should be done by tomorrow and the ASIC team should have sample chips by Friday, or at least that's when they are expecting them (about 40) so they can run tests and make sure everything is good. Packaging should be done around the same time and shipped off to Chicago after that.

So does this mean BFL has not done any real tests on their ASIC chips?

And you all thought they were going to ship soon after getting their chips...

Pray to the ASIC Gods that the chips are 100% perfectly working.

Big surprise  

You know full well they had working prototype chips months ago, back when they were using the QFN package. They even did the testing you're so intent on bitching about. I have full confidence in their chip design.

[PuertoLibre]

Hey now, don't shoot the messenger = )

I guess we shall see what happens, I'm glad I was lucky enough to grab that screen of him giving us an update.

Also, this happened too - http://mineforeman.com/2013/02/20/kanoi-from-the-cgminer-crew-posts-photos-of-his-butterfly-labs-visit/

Excerpts:

[08:22 PM]   SLok : pictures all seen before, so the latest 18th bumping is off, but no actual word on what's next? No shipping in febr. then, not in first half march too. If you are not in batch one, expect april

[08:23 PM]   SLok : you think next batches of wafers will be treated different?

[08:24 PM]   SLok : by the bumping/packaging guys?

[11:20 PM]   BFL_Josh : I got more info today. The bumping should be done by tomorrow and the ASIC team should have sample chips by Friday, or at least that's when they are expecting them (about 40) so they can run tests and make sure everything is good. Packaging should be done around the same time and shipped off to Chicago after that.
[11:21 PM]   Bowjob : Josh, will you be shipping by March or not?
[11:21 PM]   KrLos : so, maybe the shippings start after 1 march?

Posted under the FAIR USE act for the purposes of discussion and education.

[dresdenreader]

Basic gist was that they should have the boards at their KC facility by next week sometime. This is after their test chips get to the CA facility hopefully by Friday (possibly Sat or Mon).

He made a post on another thread stating that they hope to clear out the current backlog by end of April early May. However, just an estimate. I think he's done making hard estimates for now since it's burned him so much.

He legitimately sounds like a tired dude just trying to get the product out the door. But we will see what happens in the coming weeks and what not.

Still... only 3 Avalons confirmed in the wild.

The REAL winner right now is ASICminer. Making bank right now, but I just can't imagine their power costs.

[kaerf]

FYI, here's some info regarding the bumping too. not sure if this info was posted elsewhere

08:04 <@kanoi> Well I got the full Bump story today
08:04 -!- BlackPrapor [~kvirc@94.41.201.80.dynamic.ufanet.ru] has joined #cgminer
08:04 < con_> hmm I worry when the highest bidder has zero purchases to date...
08:04 <@kanoi> anyone wanna hear?
08:04 < con_> of course ^_^
08:04 < RommelVR> ha
08:05 <@kanoi> Firslty they did indeed offer them a lot of extra money to do it fast ...
08:05 <@kanoi> But they couldn't do it fast
08:05 < con_> lol
08:05 <@kanoi> Idiots doing the bumping
08:05 < RommelVR> fwiw (http://198.245.60.111/Pix/WorkRoom1.jpg <- NO PHONE ZONE ... naughty)
08:06 <@kanoi> Sonny's guess was that the tech guys got none of the extra so they wouldn't do it so it was probably the managers' (plural) fault
08:06 < con_> I assume bumping cannot be done in house for just one device or something?
08:06 <@kanoi> no
08:07 < con_> hmm might be a big night here, grabbing a coffee brb
08:07 < con_> feh coffee machine's cold, just turned it to heat
08:07 <@kanoi> anywya - since there is no blank wafer to test and align the first bump, they gotta be REAL careful of lose 1100 chips
08:07 < con_> what kind of department is this? Everyone knows anaesthetists need coffee asap
08:08 < con_> yowzer
08:08 <@kanoi>
08:08 < con_> that would fuck everything over
08:08 <@kanoi> well only one wafer - but yeah would be a small setback

[dresdenreader]

Ouch.

Anyway, for anyone interested, this is the easiest way to see all of his updates.

https://forums.butterflylabs.com/vbshout.php?message=&username=BFL_Josh&hours=&from%5Bmonth%5D=2&from%5Bday%5D=20&from%5Byear%5D=2013&end%5Bmonth%5D=2&end%5Bday%5D=21&end%5Byear%5D=2013&filter%5Bshout%5D=1&chatroomid=0&orderby=DESC&perpage=100&s=&do=archive&instanceid=1

[repentance]

At this stage BFL would be better off saying "mid-March" and pleasantly surprising people by being ready to ship early than giving another overly optimistic "worst case scenario" and failing to meet it.  22 February was supposed to be the "worst case scenario".

I'm also curious about how much their units weigh because $88 for international shipping sounds way too low.

[Puppet]

You know full well they had working prototype chips months ago, back when they were using the QFN package. They even did the testing you're so intent on bitching about. I have full confidence in their chip design.

There is only so much you can test using a different package, particularly if its true those chips destroyed themselves from heat.

Fact is they they are yet to do the first tests of BGA chips,  and even assuming they managed to do basic functional verification on the QFN chips, there are a millions things different now. A package change is no minor tweak, it changes the interconnects, signal integrity, the package and ball alignment have to be checked,   it completely voids any thermal or physical stress tests, the PCBs are new, its probably safe to say they have will have to redo chip characterization if they ever did any previously, they can only guess how the underfill will cope (assuming they used any, if not that might be a serious reliability problem). You know, these are the kind of things that even highly experienced companies like nVidia or AMD get wrong occasionally. For instance here:
http://www.theinquirer.net/inquirer/news/1050052/nvidia-chips-underfill

Just small snippet from the above link:

The next theory is slightly more plausible - that Nvidia didn't have time to properly test. A heat cycle test of packaging material takes about three months to do, and you can't really rush it.
The Inquirer (http://s.tt/14c7t)

 Ill grant that a bitcoin asic is substantially less complicated than a modern GPU but still..

[PuertoLibre]

At this stage BFL would be better off saying "mid-March" and pleasantly surprising people by being ready to ship early than giving another overly optimistic "worst case scenario" and failing to meet it.  22 February was supposed to be the "worst case scenario".

I'm also curious about how much their units weigh because $88 for international shipping sounds way too low.

Hopefully it is not by USPS or EMS. USPS works with EMS last I checked and USPS recently jacked their prices sky high for international shipping. (EMS = Express Mail Service http://www.ems.com.cn/english.html)

[repentance]

At this stage BFL would be better off saying "mid-March" and pleasantly surprising people by being ready to ship early than giving another overly optimistic "worst case scenario" and failing to meet it.  22 February was supposed to be the "worst case scenario".

I'm also curious about how much their units weigh because $88 for international shipping sounds way too low.

Hopefully it is not by USPS or EMS. USPS works with EMS last I checked and USPS recently jacked their prices sky high for international shipping. (EMS = Express Mail Service)

I plugged a 6 pound package into DHL's calculator.  For 3 day express shipping - leaving Kansas City on a Monday and landing in Sydney on Thursday the same week, it was $149 if organised online (this doesn't take into account any customs delays - they only guarantee getting it to Sydney in that time-frame).  Once things are cleared through customs and handed off to local carriers, delivery is usually pretty quick.  You can end up having to pay for redelivery, though, because the tracking information isn't reliable with some carriers and they attempt delivery before the date shown on the tracking info.

I know there've been a lot of customs problems with EMS packages coming from China to Australia recently - don't know if packages from the US are affected.

Starting packing the same day as testing starts is a huge risk, but I suspect they wouldn't be able to get another slot with the packaging facility any time soon if they hold off until the chips are tested.

[PuertoLibre]

At this stage BFL would be better off saying "mid-March" and pleasantly surprising people by being ready to ship early than giving another overly optimistic "worst case scenario" and failing to meet it.  22 February was supposed to be the "worst case scenario".

I'm also curious about how much their units weigh because $88 for international shipping sounds way too low.

Hopefully it is not by USPS or EMS. USPS works with EMS last I checked and USPS recently jacked their prices sky high for international shipping. (EMS = Express Mail Service)

I plugged a 6 pound package into DHL's calculator.  For 3 day express shipping - leaving Kansas City on a Monday and landing in Sydney on Thursday the same week, it was $149 if organised online (this doesn't take into account any customs delays - they only guarantee getting it to Sydney in that time-frame).  Once things are cleared through customs and handed off to local carriers, delivery is usually pretty quick.  You can end up having to pay for redelivery, though, because the tracking information isn't reliable with some carriers and they attempt delivery before the date shown on the tracking info.

I know there've been a lot of customs problems with EMS packages coming from China to Australia recently - don't know if packages from the US are affected.

How much is it for a 33 pounds package via DHL vs EMS?

Edit: Origin: China

[repentance]

How much is it for a 33 pounds package via DHL vs EMS?

Edit: Origin: China

Here's a link to the DHL calculator.  I can't find one for EMS.

http://dct.dhl.com/input.jsp?langId=en&originCCId=CN

[hardcore-fs]

you are not taking into account the fact thats If you pull a rep. in and offer them a long term contract, prices rapidly fall
Usually by over 50%.

The reason being that they can bulk load the goods on pallets , then split shipments at a Shippers terminal.
For example if you have 50 units going to Europe, by palleting then splitting the load in Europe, the shipping costs are WAY lower than sending single products.
There are all sorts of cost savings that can be made for a 'contract', and it is nothing like typing a few numbers into a web page.

[repentance]

you are not taking into account the fact thats If you pull a rep. in and offer them a long term contract, prices rapidly fall
Usually by over 50%.

The reason being that they can bulk load the goods on pallets , then split shipments at a Shippers terminal.
For example if you have 50 units going to Europe, by palleting then splitting the load in Europe, the shipping costs are WAY lower than sending single products.
There are all sorts of cost savings that can be made for a 'contract', and it is nothing like typing a few numbers into a web page.

This is true but it generally only works well if you're shipping in reasonable sized batches on a regular basis.  It's useless when you're dribbling stuff out the door day by day.  Here at least, you generally need to commit to a certain volume per month (usually by either weight and/or cubic metres as the agents essentially purchase the capacity in advance) and the shipment times can be less flexible than dealing with the freight companies directly.

BFL would have been up shit creek if they'd signed shipping contracts based on any of their projected shipping dates.  Even now, they don't really know their assembly capacity, when subsequent batches will really be ready for shipping or what demand will be once back orders are finally filled.

[Frizz23]

Still... only 3 Avalons confirmed in the wild.

I hope we don't say the same thing about BFL in a couple of month.

Is there an estimate how many units BFL can produce per day? If we take Avalons output as a ballpark figure we are all fucked ...

[repentance]

Still... only 3 Avalons confirmed in the wild.

I hope we don't say the same thing about BFL in a couple of month.

Is there an estimate how many units BFL can produce per day? If we take Avalons output as a ballpark figure we are all fucked ...

The reflow oven in the pictures posted can produce 700 boards per day according to the manufacturer.  Obviously there are going to be different lead times for the components.  They probably don't have a realistic idea of their assembly capacity at the moment.  

Once they're caught up on their back orders, they may very well be able to keep pace with demand.  Trying to complete back orders in-house mightn't be the wisest choice if it means that by the time the current back orders are cleared (estimated at May-June), they'll still be running two months behind on order fulfilment but BFL doesn't seem inclined to suspend new pre-orders until they catch up so that come June they can go to a turnaround time of under week.

I feel like I'm the only BFL who has some sympathy for Josh.  He's just the messenger and people's ire should be directed at those who are actually in control of the project - but those people are largely invisible.

[Puppet]

I feel like I'm the only BFL who has some sympathy for Josh.  He's just the messenger

No he isnt. He is COO.  And he better start acting like one rather than acting as a messenger/apologist.
I wont hold him personally accountable for all the previous missed deadlines before he got hired,  but he is responsible for making and meeting them now.

[meowmeowbrowncow]

I feel like I'm the only BFL who has some sympathy for Josh.  He's just the messenger

No he isnt. He is COO.  And he better start acting like one rather than acting as a messenger/apologist.
I wont hold him personally accountable for all the previous missed deadlines before he got hired,  but he is responsible for making and meeting them now.

Something that people conveniently forget, or use selectively, is that everything is personal.


One has a responsibility in what they participate in. 


To deny this is to deny the notion of integrity.