Finally photos of real BFL Single SC and Jalapeno ASICs

[Starlightbreaker]

No, that just means that they don't have a prototype of this board running, it does not mean they are "fakes"

A fake is something intended to fool. An engineering sample is a more likely possibility.

Engineering samples are by definition an actual, mostly-working version of the chip, possibly with various issues and really poor yields, but functional enough to test with. BFL have said they don't have any chips like that yet, so these pretty much have to be fakes (though I believe the more polite term is "mock-up").

Oh my.

I can't quite believe they are ordering up 20,000 chips, and counting on them to all work right, without having first received such an engineering sample, to verify that it works, and to have the opportunity to make any changes before producing them in large numbers.

If they are, well, best of luck gentlemen.

You make a lot of assumptions in that sentence.

If you recall the announcement, Josh specifically mentioned the "final chip versions," so I'm going to stand well back because you are going off half-cocked, and that is dangerous to your fellow Musketeers.
(realizing some of our international friends might think this colloquialism is a porn reference, here is some background: http://en.wikipedia.org/wiki/Half-cock)

It would appear from the wording that there has been at least one other batch or test ASIC. FYI, I've not looked him up yet, but the guy they have doing the ASIC design is supposed to be pretty hot stuff.

and all those 20000 chips are guaranteed to work?

i don't know about the failure rate, but i can safely assume not all of the 20000 will be failure-free....i think.

[1714894524]

1714894524

[1714894524]

1714894524

[scrybe]

and all those 20000 chips are guaranteed to work?

i don't know about the failure rate, but i can safely assume not all of the 20000 will be failure-free....i think.

It all depends on the contract, was it for X Working Devices to spec Y, or was it just X Devices with a penalty for excessive defects beyond a certain percentage? We don't know, but it really does not matter.

They have stated "This is enough for the pre-orders and then some", this means that even if there are faulty ASIC's they should have enough for the current orders.

[Frequency]

and all those 20000 chips are guaranteed to work?

i don't know about the failure rate, but i can safely assume not all of the 20000 will be failure-free....i think.

It all depends on the contract, was it for X Working Devices to spec Y, or was it just X Devices with a penalty for excessive defects beyond a certain percentage? We don't know, but it really does not matter.

They have stated "This is enough for the pre-orders and then some", this means that even if there are faulty ASIC's they should have enough for the current orders.

quality from China sucks really hard sorry to say so.

If send specific specs to China to make a batch of bolds metric screws specific measurements,
order 100.000 you be sure that u can thro away 20%..  but he the remaining 80.000 is still cheaper to make over there then produce them in Europe or USA. Quality is no there middle name...

[makomk]

If you recall the announcement, Josh specifically mentioned the "final chip versions," so I'm going to stand well back because you are going off half-cocked, and that is dangerous to your fellow Musketeers.
(realizing some of our international friends might think this colloquialism is a porn reference, here is some background: http://en.wikipedia.org/wiki/Half-cock)

It would appear from the wording that there has been at least one other batch or test ASIC. FYI, I've not looked him up yet, but the guy they have doing the ASIC design is supposed to be pretty hot stuff.

His wording makes it sound like that, but someone asked whether they had any prototypes that were actually hashing at all and Josh said no:

Inaba, I guess you didn't see my question yesterday. Does BFL have a working (hashing) prototype ASIC miner?

No, we are waiting on our ASIC chips right now, as I've stated in a number of other places, though it's understandable if you have missed the posts, since they are spread out everywhere.

Admittedly that was a week ago, but he was using similarly confusing language back then too. For instance, he said: "We are waiting on the bulk chips from the foundry, that is really the major delay right now" as though there were some previous non-bulk chips, which apparently don't actually exist.

[scrybe]

His wording makes it sound like that, but someone asked whether they had any prototypes that were actually hashing at all and Josh said no:

Inaba, I guess you didn't see my question yesterday. Does BFL have a working (hashing) prototype ASIC miner?

No, we are waiting on our ASIC chips right now, as I've stated in a number of other places, though it's understandable if you have missed the posts, since they are spread out everywhere.

Admittedly that was a week ago, but he was using similarly confusing language back then too. For instance, he said: "We are waiting on the bulk chips from the foundry, that is really the major delay right now" as though there were some previous non-bulk chips, which apparently don't actually exist.

You have to look very carefully at the question and how it is asked. There are 2 questions here. Did you create and test a prototype? Is this prototype still running on MainNet (or has it ever?) The only way Josh would answer True to the compound question would be if they still had a prototype running. If they had a prototype that is no longer running then his answer would be correct, even though it was not what we were looking for.

Given the number of threads accusing BFL of mining on MainNet I can't be too annoyed that he wanted to put a nail in that half of the coffin.

At this point it is too late for much, so I'm not going to worry about it.

[kwoody]

I think it's in everyone's interest that no ASIC hits the market before the block reward halving. Less volatility in the supply of BTC that way. Predicting an early December release from both BFL and Tom(caplepair).

[scrybe]

I think it's in everyone's interest that no ASIC hits the market before the block reward halving. Less volatility in the supply of BTC that way. Predicting an early December release from both BFL and Tom(caplepair).

You are no fun, we need like 2 weeks of kaboom! to have stories to tell our grandkids about, and then get on with the craziness of ASIC adoption. Blocktimes are likely to go under 5 minutes anyway, softening the impact of the halving for 2-4 weeks:

"Date & Time   Block Time      Block Value

11/28/12 0:00   5.338700753   50
12/5/12 11:22   4.520922777   25
12/11/12 19:17   7.249945409   25
12/21/12 22:52   8.167824744   25
1/2/13 9:19      8.433939204   25
1/14/13 4:42   8.357994714   25
1/25/13 21:31   8.32945169      25
2/6/13 13:23   8.671654753   25


This is from a forecasting model based on the actual preorder numbers from Tom, ASIC lot Qty info from BFL, ASICMINER based on announcements in January, along with 2 waves of Avalons. Shipments of preorders are spread out over this whole section of difficulties for BFL, intial batch shipping within 2 weeks for Tom (WAY different sizes, does not reflect badly on BFL) and smaller follow-on batches every 2 weeks.