Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Odd 970 Photos
- Thread starter MacRumors
- Start date
- Sort by reaction score
Very Nice
These pics shouldn't be derided as trash! They are very tastefully done! And I'm sure the accompanying article is fascinating, too.
These pics shouldn't be derided as trash! They are very tastefully done! And I'm sure the accompanying article is fascinating, too.
Re: Re: How about some G4 on G4 action?
It will be a matter of IF up until the time Apple announces their plans to use the 970.
It will be a matter of IF up until the time Apple announces their plans to use the 970.
Re: confirmed 970??
Unless the writer of this article works for IBM then anything said about Apple using the processor is pure speculation. The article suggests that Apple would be using the 970 since OS X deserves a more powerful processor. Until Apple says it, it is just a rumor.
Unless the writer of this article works for IBM then anything said about Apple using the processor is pure speculation. The article suggests that Apple would be using the 970 since OS X deserves a more powerful processor. Until Apple says it, it is just a rumor.
chewbaccapits
macrumors 6502a
Feathers
I would take the feathers as an insider joke on one of the development code names for 970 (ultralight).
I compared the photos with the feathers to the 970 die photo available from the same gallery.
The aspect ratio of the chip pictured with the feathers appears to match the aspect ratio of the die photo of the 970. The die photo is likely of a chip pulled from the line before the upper metal layers were deposited. This gives a picture which highlights the floorplanning.
Chips fabricated all the way to the C4s (like the one pictured with the feather) are pretty boring and uniform looking because most of what you're seeing are the C4s (solder balls), power grid, and clock grid.
As an aside, it is not clear to me whether that "die photo" is an actual photo. It looks suspiciously regular. Real die photos have variations in color between different instances of the same circuit because of the different light paths and interference patterns.
The "IBM 64-bit PowerPC 970 microprocessor chip die. (#88)" looks too regular. Notice how the different sub-blocks of the caches are all exactly idential to each other. This reminds me of the fake die photo plots we use internally for presentations, posters, etc. They are much more satisfying than an actual die photo because they reveal more of the internal structure...it's easier to point to "what I designed".
C4 = Controlled Collapse Chip Connection, introduced in products in 1964.
I would take the feathers as an insider joke on one of the development code names for 970 (ultralight).
I compared the photos with the feathers to the 970 die photo available from the same gallery.
The aspect ratio of the chip pictured with the feathers appears to match the aspect ratio of the die photo of the 970. The die photo is likely of a chip pulled from the line before the upper metal layers were deposited. This gives a picture which highlights the floorplanning.
Chips fabricated all the way to the C4s (like the one pictured with the feather) are pretty boring and uniform looking because most of what you're seeing are the C4s (solder balls), power grid, and clock grid.
As an aside, it is not clear to me whether that "die photo" is an actual photo. It looks suspiciously regular. Real die photos have variations in color between different instances of the same circuit because of the different light paths and interference patterns.
The "IBM 64-bit PowerPC 970 microprocessor chip die. (#88)" looks too regular. Notice how the different sub-blocks of the caches are all exactly idential to each other. This reminds me of the fake die photo plots we use internally for presentations, posters, etc. They are much more satisfying than an actual die photo because they reveal more of the internal structure...it's easier to point to "what I designed".
C4 = Controlled Collapse Chip Connection, introduced in products in 1964.
Feathers = substrate?
Actually, I'm wondering if the fact that the chip is among feathers might not somehow be a reference to the fact that compressed feathers are being suggested by some as a replacement for the silicon wafers currently used in semiconductor manufacturing.
Check out the details at the following URLs:
https://www.ccm.udel.edu/news/July02News/July02News-ACRES1.htm
http://www.acm.org/technews/articles/2002-4/0708m.html
http://www.informationweek.com/story/IWK20020802S0025
http://www.acm.org/technews/articles/2002-4/0708m.html
Most of these articles are dated July of last year, so it isn't all that brand new. Even the people over at IBM could've possibly heard about it by now. 🙂
--Patrick
Actually, I'm wondering if the fact that the chip is among feathers might not somehow be a reference to the fact that compressed feathers are being suggested by some as a replacement for the silicon wafers currently used in semiconductor manufacturing.
Check out the details at the following URLs:
https://www.ccm.udel.edu/news/July02News/July02News-ACRES1.htm
http://www.acm.org/technews/articles/2002-4/0708m.html
http://www.informationweek.com/story/IWK20020802S0025
http://www.acm.org/technews/articles/2002-4/0708m.html
Most of these articles are dated July of last year, so it isn't all that brand new. Even the people over at IBM could've possibly heard about it by now. 🙂
--Patrick