Topic: BFL ready to ship? (Read 10390 times)

1) This is simply wrong.  When converting from QFN to FCBGA, the upper layer of metal on the silicon is changed to a miniature version of the BGA package.  Proof of this is the photos of the wire bonding test die.  The attachment points are (now) in an array, not along the edge.  Note that the die can have typically more or sometimes less connections than the final BGA package.  The BGA package is simply a very tiny multi-layer PCB that translates the chip balls/bumps to the package ball grid on the bottom.

What, they didn't apply the glue that holds the metal caps in place evenly?
 

Enigma pointed to strange looking soldering points, but:
- they surely use mostly the edge pins of the chip (because of quick QFN to FCBGA conversion), leaving  most of the middle pins unused;
- Enigma used a photo of a board which was put through the oven without chips, not the one just prepared to place them;
- the photo was blurry and unfocused, so Enigma's conclusions are plain guesswork.

What, they didn't apply the glue that holds the metal caps in place evenly?
 

Good god

It won't take much either.  Just the normal day to night swings will do it.  I have seen 100s of thousands of underfilled chips over 10 process generations, many with underfill defects, and none of them looked as bad as these.

The black blotch shown on greyhawk's zoom is a scrap part.  That is underfill, it will prevent proper cooling and is impossible to sand off without cracking the brittle silicon.

I can't comment on the validity of his statements, but ~1000/day isn't unheard of.

Yes, because that completely invalidates the garbage underfill job, and the chipped dies that Greyhawk pointed out.

Sprinkler you say? Good, ´cos Entropy got me worried for a second. After all he is a specialist payed with 1000$/day to spot this kind of problems.

I recognized it as a nano-sprinkler straight away. Genius!

Are you guys sure this black spec is on the chip itself? Looks to me like something on the ceiling reflected on the chip....

Are you guys sure this black spec is on the chip itself? Looks to me like something on the ceiling reflected on the chip....

That was my reasoning which is why I did not arrow it.

Are you guys sure this black spec is on the chip itself? Looks to me like something on the ceiling reflected on the chip....

Are you guys sure this black spec is on the chip itself? Looks to me like something on the ceiling reflected on the chip....

It won't take much either.  Just the normal day to night swings will do it.  I have seen 100s of thousands of underfilled chips over 10 process generations, many with underfill defects, and none of them looked as bad as these.
The black blotch shown on greyhawk's zoom is a scrap part.  That is underfill, it will prevent proper cooling and is impossible to sand off without cracking the brittle silicon.

I really don't think Josh is as aware of these types of situations as you are. Why don't you shoot them a quick message so he gets the idea. It would be fun to see him screaming at the packagers.

Thanks for asking.  I have a few simple points.

About BFL
1.  You are pre-announcing this product before delivering your previously pre-announced product
2.  This is your second claim of delivering a product with ASIC hashing.  Your first was debunked as an FPGA as soon as product shipped.
3.  You seem to be using customer funds for purchases of product for development and general expenses.  If I am wrong about that please post a notarized statement from your escrow agent and I will apologize.

About business in general
1.  Paying in advance for development is an investment.  It is generally compensated with intellectual property, equity or loan interest.
2. Honest, viable businesses become insolvent all the time for lack of cashflow.  It is the reason for the Chapter 11 bankruptcy procedures.
3. Payments in advance for products become unsecured debts in a bankruptcy proceeding.  Unsecured debtors are among the lowest priority for recovery in these procedures and generally get nothing.  Even if the material is sitting in the shipping bay with a label on it, it isn't yours under bankruptcy.
4.  For reason #3 above it is unethical, and a violation of consumer protection laws in many places to sell product in advance of availability.  For the same reason, Visa and Mastercard require shipment before a sale can be charged.
5. Companies at risk of insolvency often make very generous guarantee offers.  A guarantee has no revenue cost in the present, and has no value in bankruptcy either.

I can't see ever being a customer of yours given your business practices.  And your question has given me the opportunity to  nicely summarize why people should use caution.  So I will leave your thread to your fans and investors.

It won't take much either.  Just the normal day to night swings will do it.  I have seen 100s of thousands of underfilled chips over 10 process generations, many with underfill defects, and none of them looked as bad as these.

The black blotch shown on greyhawk's zoom is a scrap part.  That is underfill, it will prevent proper cooling and is impossible to sand off without cracking the brittle silicon.

Join my Posse.

[Image removed...]

I just don't know what the Posse is about yet. Perhaps you should start selling PL merchandise? Make a whole franchise out of it.

I might be more popular than the Pope. Well, ex-pope.

It's a reliability problem.  Thermal cycling of the chips will cause the underfill to delaminate at the void sites and break solder joints.  So you put it all together and everything looks sweet, then a month or 2 down the road the chip goes completely flakely on you and eventually fails completely.
The only way to know what is going on is to inspect some of the die using CSAM.  It's a non-destructive process and can be done with a sample of packages so it's not the end of the world.  Unless they are all full of voids...

I like this picture, where is it come from?

Pics of chips != ready to ship

There, you can see it. Josh tweeted a pictures of the packaged chips a few minutes ago (source: https://twitter.com/ButterflyLabs/status/310549500311789570 )

I like this picture, where is it come from?

Cost aside, wouldn't that cause further delay? They could inspect them during the built process. If a unit requires 8 chips, plug in 7 that look okay and 1 (?). This way two issues could be solved with one (insert noun, but not stone).

Yes, I can really tell.  A properly underfilled chip will have a smooth line where it was dispensed on one side, and an equally smooth line on the other side where the epoxy flowed through. The clumping and discontinuities in the images make it likely there are voids.  To be certain you would have to inspect the die with CSAM.  The epoxy should develop a clean curve up the sidewall of the die as well - this shows good adhesion to the silicon, and ensures that the epoxy doesn't interfere with the heat sink mounting.

It's really ugly. If I was in Josh's shoes I'd be screaming mad, checking every chip to make sure the underfill isn't higher than the die back side, and demanding CSAM inspections from the assembly house to prove the packages are free of voids.

I would also like to know this.

Zoom up on the full sized image and look at the edge of each die.  That is the ugliest underfill job I have ever seen on chips.  They are going to have nasty voiding on the C4 bumps and rel problems out the wazoo.

There, you can see it. Josh tweeted a pictures of the packaged chips a few minutes ago (source: https://twitter.com/ButterflyLabs/status/310549500311789570 )

I think I am honored. Though I will have to check with my therapist to make sure if I am or am not honored?

I believe when I see it...

So many BFl cheerleaders in here. Or PL haters lol...either way i'm too here to expose BFL for their bullshit.

Its true Puerto. I'm a BFL-hater and even I think you're just being a dick.  A quite obviously insecure one at that.

Yes, obviously you don't care since you took the time to respond [/sarcasm]
Never said he didn't say it, but it was an offhand comment in the middle of intensive chat about gaming consoles which you conveniently ignore.  I'm not a BFL defender, I'm a PuertoLibre criticizer. You're just here to be a dick and get attention-that's the problem I have.  There are a lot here like you, I know, but you're just the most annoying and obvious in the threads that interest me.

Yes, obviously you don't care since you took the time to respond [/sarcasm]

Never said he didn't say it,

but it was an offhand comment in the middle of intensive chat about gaming consoles which you conveniently ignore.

I'm not a BFL defender, I'm a PuertoLibre criticizer.

You're just here to be a dick and get attention-that's the problem I have.  There are a lot here like you, I know, but you're just the most annoying and obvious in the threads that interest me.

TCP RST, you can look it up.

It is at about 3/7/2013 at 11 o'clock.

Look through the shouts archive. You'll see the statements refereed to there.

---------------------

As for your comments, I don't really care what you think. Truth hurts I guess.

What I'm interested now is that 28nm helveticoin asic offer

Now we know that avalon can basically follow their schedule with almost no delay, that is respectable speed. If they managed to quickly push the 28nm product out in same timeframe (4 month), BFL will have a hard time. If I understand it correct, there will be at least 3 vendor building those ASIC machines and put at least 1000 TH on network by summer/autumn

TCP RST, you can look it up.

It is at about 3/7/2013 at 11 o'clock.

Look through the shouts archive. You'll see the statements refereed to there.

---------------------

As for your comments, I don't really care what you think. Truth hurts I guess.

Update:



https://forums.butterflylabs.com/bfl-forum-miscellaneous/1221-there-problems-asic-chips-3.html#post17536

Possible short changing of customers in the near future.

Let BFL know ( [email protected] ) how you feel about such "a drastic" option. As this would reduce your overclocking potential in the long run and reduce the long term value...it is bad news bears.

That'd be awful .  The least we deserve as pre-order customers who've waited the better part of a year is a possible overclock.  Also, this would put Jalepeno customers at an advantage, as it is just one chip.  I wonder if BFL would actually use chip-sized spacers and short the pre-order customers of Singles, Mingles, and Mini-Rigs?   I could understand doing this on later non pre-order sales when HSFs and specs can be redesigned and finalized, but for pre-order customers that paid last June for an order that should have shipped in October?  I really hope this isn't the case.

I wouldn't say "guessing". It's great marketing getting people to pay for something in advance where getting it first counts so much, even if it's not likely to be available for another 6 months after you claim.

pretty dishonest:(

i keep winning bets.. first one i won was with p4man back in 2011 lol...

i wonder where that kid ended up...

it seems bfl could need a month or more if all goes well, why do they keep guessing wrong on their time table?

it seems bfl could need a month or more if all goes well, why do they keep guessing wrong on their time table?

and your question is ridiculously off-topic.

The chips are being packaged as we speak. I don't know if I'll have an update on whether or not that's completed tonight, but I should have an update sometime in the morning/early afternoon tomorrow at the latest. If I get an update today, I will, of course, post it. But the packaging house did confirm they have the chips and are packaging them today.

Did bitcoin ever experience a crash?

Forgot to mention the time it takes to place on the following labels:

Is this true or is this bull?

http://www.youtube.com/watch?v=UX75tiViEjA

Did bitcoin ever experience a crash?

No, that was when a hacker broke into Mt Gox and brought the market down to $0.01. Those fake trades were rolled back. The market did crash from it's bubble high to a couple dollars but nothing like that.

Is this true or is this bull?

http://www.youtube.com/watch?v=UX75tiViEjA

Did bitcoin ever experience a crash?

I got some bets going on this

I recon Difficulty will hit 6,500,000+ before a BFL ASIC product hits costomers.
BFL will Deliver ASIC before bitcoin difficulty reaches 6,500,000

I recon BFL will deliver before the network reaches 50TH.
The BTC Mining hashrate will reach 50TH before BFL deliver an ASIC


They're both nearly too close to call I bet

All of you clearly don't understand ship dates.  I believe the BFL products will ship by February 47, 2013 or maybe Feb 52 at the latest. 

As you can see their schedule is solid, it does not change. If a man of word said "February" then neigher March nor other month can make him take his words back.

Looks like "schedule" changed. BFL fan boys read to give in to the fact that BFL is full of BS? 

Everybody worship BFL, they have clear terms of pre-order!

http://products.butterflylabs.com/featured-products/60gh-bitcoin-miner.html


Oh, shit. Its already March 2013. But I think Inaba can easily fix this issue by using some insults here and there.

They took 1 of the 6 wafers from the bumping facility to wire it up for testing.  
Q: Why?
A: They did this because bumping was taking awhile and they wanted to be able to give the fab the go ahead to put the final layers on the other 6 wafers in waiting.
Q: What happened?
A: They say the test went fine, everything is okay.
Q: So did they tell the fab to finish putting the final layers on the 6 wafers in waiting?
A: No...
Q: Why?
A:  They said they're going to wait to do final testing of the first 5 wafers leaving the bumping facility going onto packaging before initiating the final layer production at the fab.

What does this tell you?
Either they're idiots and just threw away 1/6th of their first run for no reason or... something didn't go as they had planned with the wire bond testing.

No, everything is going just fine.

Maybe you're confusing the packaging of finished miners with the packaging of the chip dies inside FCBGA chip packages, which is what's being done now.

There are still a lot of steps before they can ship a finished product.

I just saw this at the BFL forums....

Ready to ship? LOL, no. Maybe reading to start assembling and testing a prototype.

Cliff notes: will ship by April-May, MAYBE, if everything goes according to plan.

I just saw this at the BFL forums....

Maybe two weeks? We are waiting for the test lab to issue the test report.

With the bump in power requirements on the MR and the new screen, we had to make changes, although the new screen is already certified.  We are doing all the devices at once, since they all share the same board.

Bumping is done and chips should be getting shipped off to the packager Saturday.