Topic: BFL Batch #1 Order Cut Off estimate (Read 3257 times)

Order #'s start at 1656 as the first order #  

Don't forget,alot,I mean alot of order #'s are unfilled.People have stated that it took as many as 3 tries to get an order placed,it took me 2 tries,so deduct my 2 & as as many as 1 per order(or .5) to get an idea of "fulfilled" orders

Only pre-orders from the first month are included in the 1/3 shipping plan.  1/3 goes to upgrades, 1/3 goes to new orders (ie, people who aren't upgrading) and the remaining 1/3 is randomly chosen from orders for both.  It's presumed that the 1/3 to upgrades and new orders will be distributed by order date.  The remaining third will still go to people who ordered in the first month but it will effectively be a lottery. How many of each unit they make out of the first batch is obviously going to affect how many people's orders are fulfilled under the 1/3 shipping plan.

At this stage, no-one is expecting that all orders from the first month will be filled by the 1/3 shipping plan - there'd be no need for the random 1/3 (which was meant to prevent a concentration of hashing power in the hands of those who ordered big and early) at all if there were enough units in the initial batch to fill all pre-orders placed during the first month.

Josh said a while back that he expects no matter how they choose to handle shipping of the initial batch, there are going to be some unhappy customers.  The 1/3 plan was formulated when they expected to receive more chips in the initial batch.

The best place for information on what's happening with any of this - the 1/3 shipping plan, shipping costs, taxes/duty customers may need to pay to receive their units - remains the BFL forums themselves.  These questions are being discussed constantly over there and none of the threads are so long that it's onerous to read them from start to finish.

Also, order numbers mightn't be useful in trying to estimate a cut-off.  It looks like the order numbers may have been segmented for the different types of units.

https://forums.butterflylabs.com/bfl-forum-miscellaneous/952-new-account-page-website-shows-your-orders-3.html#post13597

I dont get it, How are they supposed to have a cut off and a 1/3 shipping plan? Are they shipping 1/3 from all orders or saying the first 3000 orders have a chance of being in the 1/3 shipping plan.

I am getting frustrated about the lack of info from BFL, Where has josh gone? Where are we? Where am I? and for the love of doughnuts where are my chocolate sprinkles?

6000 Chips vs What if they build only one type of unit:

6000/200 = 30 Mini-rigs
6000/8 = 750 SC Singles
6000/4 = 1500 Little Singles
6000/1 = Jalapenos

I guess many BFL Minirig customers may have to wait until there are more chips.

If you're in the first month, I'd say it's a pretty good bet you're getting a unit!

Well...I made a quick excel sheet with data gathered from the preorder thread, specifically from June through the end of August. Totals come out to:
Jalapeno - 157
Little Single - 18
SC Single - 228
Mini Rig - 21

Multiply in the chip counts per device (speculated, and noted above in the thread) and you get around 6253 chips. There are tons that go unreported, but given Josh's statement that they're break into the August orders, my guess is about 50% more orders than we've seen so far...bringing us to 9380 chips to cover June through August.

Obviously just a rough guesstimate, but I figure it'll work for now until we get more data.

Remember, I think awhile ago it was confirmed that the ASIC orders start at about Order #1500-ish.  True that 6K chips does not mean 6K Jalapenos.

I too was just trying to figure out this math last night and wondering how many of each unit would potentially have been orderd, along with all the false order #s (just due to how their ordering system would increment). 

While we wait and see, this #6000s range sounds good to me.  Hope to be in the first batch

Very true! Forgot about that. Keeping my fingers crossed

Let's hope you're right...seeing as I have my first order placed on the 16th and at 5467



And there aren't 6,000 orders in the first batch...there's only a few hundred. Just because you have an order number over 6000 doesn't mean there are 5999 orders before you...

Umm.. have to burst some bubbles here folks.. but that was 6,000 CHIPS in the first batch..  This does not equate to 6,000 orders.  each unit has multiple chips .... 

Sigg

I hope Yogafan is right. This would be the first good ASIC-related news for me since well, ever! All three of my units would be in the first batch. 7/26, 8/7, and 8/13 were the order dates.

Don't get too happy you August people.


Somebody quoted BFL saying their order count is somewhere between 1.5 to 2x that of what's reported on the forum.

Well that sucks. My first order is Aug 27th #6386. Guess I'll just barely miss the first batch.

Judging from the dimensions in the photo on BFL's forum of a solid block aluminum heat sink, the Mini-Single seems to be a 4 chip unit, whereas the Copper heat pipe HSF will be for an 8 chip Single (different metal plate dimensions than the aluminum non-heat-pipe version).

Hrmm.. I'm wondering if they could have up to 50% of the hashing cores bad on a chip, and use those chips in the 30Ghash version by just disabling 1/2 the cores in each chip ?

( I've always wondered just how they were gonna make the 30Ghash version.. slower clocks?  fewer chips? )

Sigg

That doesn't really affect what he said though, yields will affect the number of chips they receive since their are not getting 6000 chips in the first batch, they're getting 6 wafers. I don't believe 6000 is 100% yield though, they're expecting more than 1000 per wafer (I believe 1100). They could even have more than 6000 if yields are good. They've said that if a single hashing core was nonfunctional they can disable it and increase clocks on the other cores to compensate, so baring gross errors their yields shouldn't be terrible.

Their "yield" may not be 100%, but it is unlikely the fab will package up 'dud' chips....
The foundry will ATM the chips and mark the duds I.E chips that don't meet the specs.

I'm at #25xx
If I don't get my upsized Jalapeno, I'm refunding.

I ordered a minirig upgrade on 6/25 and I'm worried I won't be in the first batch.  

There are 16 minirigs listed in the original FPGA minirig thread.  Add to this gigavps with 4 rigs and DarkIce with 6 and there are 26 known FPGA rigs.  It is almost certain all of these will be upgraded, and most minirig owners were probably on top of ordering their upgrade.  26 known trade-in minirigs will account for 5200 chips alone.  Add to this unknown minirig trade-ins ,new minirig orders, and of course singles and jalepenos.

6000 working chips is optimistic as well, their yield will not be 100%.

By my estimate, the "BEST CASE SCENARIO" cut off date for BFL's batch #1 is for orders placed on or before August 16, 2012, or order number 5515 or lower.  This will give you a ballpark idea if you will recieve your BFL ASIC from the 1st shipment of chips.

I arrived at this number from this quote by Josh on BFL forums:


and using the formula of 1 chip per Jalapeno, 4 chips for little single, and 8 chips for SC single, and 200 chips for Minirig, combined with the info from the waitlist thread found here. https://bitcointalksearch.org/topic/wait-list-bfl-sc-pre-order-information-monarch-orders-too-89685

Caveat: This estimate is the perfect scenario where every order is posted on the wait list and BFL recieves 6000 chips with no duds.

Obviously, less working chips and unreported orders will affect the cutoff order date negatively.