[Announcement] Avalon ASIC Batch #1 Ships

[PuertoLibre]

You guys/gals do not know how to use that tool properly.

You folks don't even know what a JPEG matrix or quantization is nor how the tool even detects changes in an image.

There is nothing wrong with the tool. There is only bad interpretation(s) and bad application(s) of this tool without understanding what it compares.

(I used it several years ago (2006 -2008) when the hacker who made (and introduced) this tool first came out with it. I read the "white" papers on some of his theory and how they detect changes.

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

Please don't make fools of yourselves by not knowing how to use the tool properly....

I wrote:

This photo was manipulated

This is not the original photo, you know it. I do not know what it was manipulated , It is possible that only resize. No exif also indicates manipulation.

I am not outright discouraging you.

I am just pointing out that you are using that tool in the wrong way without knowing how it's detection algorithms work and what they are intended to show.

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

You should start off with a program that can compare pixels to pixels. (not through jpeg quantification and matrix mis-matches).

I'll help you out:

http://www.abonsoft.com/pd_imagecompare.html

This is a very crappy but simple pixel level compare software. It should point out what parts are actually different if any. (without the other method which required a better understanding of what the tool actually does or how to interpret the results)

[1713936223]

1713936223

[1713936223]

1713936223

[1713936223]

1713936223

[enquirer]

Looks like there will be significant over-subscription for the second batch. How will you handle it? Some lottery or auction scheme maybe?

[PuertoLibre]

Silly question guys...Where is the logo on the left bottom chip?

Someone grab this guys avatar and post fail all over it. He's 7 tenths of the way there already....

Did you notice that the bottom row is "flipped", genius? (as in upside down)

[Dexter770221]

A3256 name of the chip.
3 unrolled cores @ 256MHz (or it is a name of SHA256 core)
Either way 3*256=768 MH/s. Or 3*333MHz=1GH/s <-more likely. 64 chips in device, 8 PCBs with 8 chips, or looking at PCB it will be more likely 16 chips onto 4 PCBs.
My BIG speculation....

[organofcorti]

I just finished a post that you all might find fun, with charts and estimates and figures that will make you jealous of Jeff Garzik, using as up to date Avalon data as possible.

9.5 ASIC choices: Avalon ships, guerilla bitcoin mining and WWJE (what will Jeff earn?)

[Smoovious]

Silly question guys...Where is the logo on the left bottom chip?

Someone grab this guys avatar and post fail all over it. He's 7 tenths of the way there already....

Did you notice that the bottom row is "flipped", genius? (as in upside down)

Hey, genius... did you notice, that even flipped, you should still see upside down text on the chip as there isn't enough free space on the right side up chips to account for all the empty space shown on the upside down chip?

-- Smoov

[DutchBrat]

Silly question guys...Where is the logo on the left bottom chip?

Someone grab this guys avatar and post fail all over it. He's 7 tenths of the way there already....

Did you notice that the bottom row is "flipped", genius? (as in upside down)

Hey, genius... did you notice, that even flipped, you should still see upside down text on the chip as there isn't enough free space on the right side up chips to account for all the empty space shown on the upside down chip?

-- Smoov

Maybe they supply sanded off chips to make life easier for BFL when they order them from Avalon

[Vicus]

Hey, genius... did you notice, that even flipped, you should still see upside down text on the chip as there isn't enough free space on the right side up chips to account for all the empty space shown on the upside down chip?

Laser engraving on chip is best viewed under certain angle.  Especially if you make foto with flash. I think, photoflash blended text with chip background.

[PuertoLibre]

The image is of poor contrast and partially out of focus. (Possible Super Macro Mode settings)

A simple color filter shows two dark lines of lettering. Not very legible. Also the filter makes the dust and lint all the more apparent. Seems like this board has been in use for a few days in a non-clean environment.

P.S. Light source seems to be coming from the top left. Likely the poor lighting is playing a role in the lettering not being apparent along the bottom.

[MrTeal]

If you've ever tried to take macro photos of PCBs, you'll know that getting the letting to show really well can actually be quite the challenge, especially on multiple chips.

@organofcorti: That's also basically a worst case scenario for Jeff, since you're showing almost all the Avalon units hitting the network by the end of January. For all we know they might not even have 100 out the door by then.

[organofcorti]

@organofcorti: That's also basically a worst case scenario for Jeff, since you're showing almost all the Avalon units hitting the network by the end of January. For all we know they might not even have 100 out the door by then.

Maybe. The assumptions I've made are listed on the blog:

Assumptions:

• 300 units will be shipped each fortnight.
• The first units of a production run will be received after three days from the start of shipping, and the remaining units from of that production run will be received within ten days of start of shipping. I'm also assuming half of the units will be received after  seven days from the start of shipping. I have nothing to base this on other than my experiences in receiving parcels, so this may be completely wrong.
• Units will arrive in a log-normally distributed fashion, with mean = log(168) hours after the first unit ships, and standard deviation of log(1.26).
• The expected delivery date for the "BFL Single SC" has been updated to the week of the 10th of February. I'm assuming the first "BFL Single SC" will arrive on the 17th of February, which means about 25 days mining from the first Avalon unit is received until the first BFL unit is received.
• The exchange rate is US$16 per btc, and Jeff's electricity costs are US$0.15/kWh
• Jeff is able to have his new Avalon ASIC hasher up and running immediately at 66Ghps and 400W, and leaves it running for the entire twenty five days. There is no downtime.

It's also possible that they're a best case scenario, and rather than 300 units per fortnight its 500 units per fortnight.

I'll write a more serious post once Avalon and BFL have shipped more than three batches each (or just Avalon, if BFL still haven't shipped) and shipping times are better known.

[DutchBrat]

@organofcorti: That's also basically a worst case scenario for Jeff, since you're showing almost all the Avalon units hitting the network by the end of January. For all we know they might not even have 100 out the door by then.

Maybe. The assumptions I've made are listed on the blog:

Assumptions:

• 300 units will be shipped each fortnight.
• The first units of a production run will be received after three days from the start of shipping, and the remaining units from of that production run will be received within ten days of start of shipping. I'm also assuming half of the units will be received after  seven days from the start of shipping. I have nothing to base this on other than my experiences in receiving parcels, so this may be completely wrong.
• Units will arrive in a log-normally distributed fashion, with mean = log(168) hours after the first unit ships, and standard deviation of log(1.26).
• The expected delivery date for the "BFL Single SC" has been updated to the week of the 10th of February. I'm assuming the first "BFL Single SC" will arrive on the 17th of February, which means about 25 days mining from the first Avalon unit is received until the first BFL unit is received.
• The exchange rate is US$16 per btc, and Jeff's electricity costs are US$0.15/kWh
• Jeff is able to have his new Avalon ASIC hasher up and running immediately at 66Ghps and 400W, and leaves it running for the entire twenty five days. There is no downtime.

It's also possible that they're a best case scenario, and rather than 300 units per fortnight its 500 units per fortnight.

I'll write a more serious post once Avalon and BFL have shipped more than three batches each (or just Avalon, if BFL still haven't shipped) and shipping times are better known.

Don't forget to include ASICMINER.... last Thursday they said it might take more than 2 weeks to get all 12 wafers online and hashing (12TH) and they have already ordered more chips to scale it up. Hopefully we will know more exact numbers in 1 1/2 weeks....

[organofcorti]

Don't forget to include ASICMINER.... last Thursday they said it might take more than 2 weeks to get all 12 wafers online and hashing (12TH) and they have already ordered more chips to scale it up. Hopefully we will know more exact numbers in 1 1/2 weeks....

Thanks for letting me know.

[MrTeal]

@organofcorti: That's also basically a worst case scenario for Jeff, since you're showing almost all the Avalon units hitting the network by the end of January. For all we know they might not even have 100 out the door by then.

Maybe. The assumptions I've made are listed on the blog:

Assumptions:

• 300 units will be shipped each fortnight.
• The first units of a production run will be received after three days from the start of shipping, and the remaining units from of that production run will be received within ten days of start of shipping. I'm also assuming half of the units will be received after  seven days from the start of shipping. I have nothing to base this on other than my experiences in receiving parcels, so this may be completely wrong.
• Units will arrive in a log-normally distributed fashion, with mean = log(168) hours after the first unit ships, and standard deviation of log(1.26).
• The expected delivery date for the "BFL Single SC" has been updated to the week of the 10th of February. I'm assuming the first "BFL Single SC" will arrive on the 17th of February, which means about 25 days mining from the first Avalon unit is received until the first BFL unit is received.
• The exchange rate is US$16 per btc, and Jeff's electricity costs are US$0.15/kWh
• Jeff is able to have his new Avalon ASIC hasher up and running immediately at 66Ghps and 400W, and leaves it running for the entire twenty five days. There is no downtime.

It's also possible that they're a best case scenario, and rather than 300 units per fortnight its 500 units per fortnight.

I'll write a more serious post once Avalon and BFL have shipped more than three batches each (or just Avalon, if BFL still haven't shipped) and shipping times are better known.

They've only sold 300 units and they've stated they won't start taking more orders until after Chinese New Years, so I don't think shipping 500 units in the next two weeks is very likely. They've also said they likely won't finish shipping before Chinese New Year but will try to get as many out as possible before then.

[DobZombie]

The image is of poor contrast and partially out of focus. (Possible Super Macro Mode settings)

A simple color filter shows two dark lines of lettering. Not very legible. Also the filter makes the dust and lint all the more apparent. Seems like this board has been in use for a few days in a non-clean environment.

P.S. Light source seems to be coming from the top left. Likely the poor lighting is playing a role in the lettering not being apparent along the bottom.

I'm still not seeing diddley squat

Someone grab this guys avatar and post fail all over it. He's 7 tenths of the way there already....

Did you notice that the bottom row is "flipped", genius? (as in upside down)

No need to be a cockhead

[Smoovious]

Hey, genius... did you notice, that even flipped, you should still see upside down text on the chip as there isn't enough free space on the right side up chips to account for all the empty space shown on the upside down chip?

Laser engraving on chip is best viewed under certain angle.  Especially if you make foto with flash. I think, photoflash blended text with chip background.

Possibly, but that doesn't look like an angle that would be an issue on...

Plus, it doesn't look flash taken at all, but done under ambient lighting. Some spots are just too dark with smooth transitions from brighter to dimmer areas.

I know that can be done with very careful flash setups, but even then, the angle just still looks off for having the printing being washed out by glare.

I'm not suggesting there is something shady with the pic. Just since someone else mentioned it, and another guy was being insulting back, calling the first poster a name, and claiming it was because the chip itself was mounted in an inverted rotation, which was why it looked like there was no printing on it, I just felt like responding in kind with my own observation.

-- Smoov

[organofcorti]

They've only sold 300 units and they've stated they won't start taking more orders until after Chinese New Years, so I don't think shipping 500 units in the next two weeks is very likely. They've also said they likely won't finish shipping before Chinese New Year but will try to get as many out as possible before then.

Ah, well then. I didn't know that. Damn that CNY.

Still, the section on guerilla bitcoin mining is valid.

[Smoovious]

The image is of poor contrast and partially out of focus. (Possible Super Macro Mode settings)

A simple color filter shows two dark lines of lettering. Not very legible. Also the filter makes the dust and lint all the more apparent. Seems like this board has been in use for a few days in a non-clean environment.

P.S. Light source seems to be coming from the top left. Likely the poor lighting is playing a role in the lettering not being apparent along the bottom.

Good enuf. I can just make out the lower portion of the L and the O in the middle of AVALON,  in the very corner of the image. Mystery solved as far as I'm concerned.

...

well... with that done, I'm bored again... what's next?

-- Smoov

[crazy_rabbit]

The image is of poor contrast and partially out of focus. (Possible Super Macro Mode settings)

A simple color filter shows two dark lines of lettering. Not very legible. Also the filter makes the dust and lint all the more apparent. Seems like this board has been in use for a few days in a non-clean environment.

P.S. Light source seems to be coming from the top left. Likely the poor lighting is playing a role in the lettering not being apparent along the bottom.

Good enuf. I can just make out the lower portion of the L and the O in the middle of AVALON,  in the very corner of the image. Mystery solved as far as I'm concerned.

...

well... with that done, I'm bored again... what's next?

-- Smoov

If it makes and different, I'm a professional photographer and this image looks absolutely normal. Shooting Macro on Circuit boards is HARD. Actually so hard that it's pretty much impossible to do in one shot (with very few exceptions).

To do it right you basically need to do macro-stacking (http://www.heliconsoft.com/heliconfocus.html) to get any sort of decent result. With a few exceptions. This looks like it's just a point and shoot. FYI: yes the bottom does look flipped but not in photoshop. Looks like the board layout is like that. Flipping stuff is easy to do it when there is no perspective involved (which there is here) which makes doing that really hard to do well. It's not impossible, but this doesn't look like that has been done.

Anyway, we will know soon enough.

[DobZombie]

The image is of poor contrast and partially out of focus. (Possible Super Macro Mode settings)

A simple color filter shows two dark lines of lettering. Not very legible. Also the filter makes the dust and lint all the more apparent. Seems like this board has been in use for a few days in a non-clean environment.

P.S. Light source seems to be coming from the top left. Likely the poor lighting is playing a role in the lettering not being apparent along the bottom.

Good enuf. I can just make out the lower portion of the L and the O in the middle of AVALON,  in the very corner of the image. Mystery solved as far as I'm concerned.

...

well... with that done, I'm bored again... what's next?

-- Smoov

Nope, still not seeing it.

I wasn't accusing anyone of being dodgy, I thought that chip may have another purpose and if there was a reason for not seeing anything.

If that is another Avalon chip and it does have something there I need my eyes tested.

Or there's nothing on there and you're all imagining things!

0.o