Every year since 2005 the 3DIC season as ended with the Research
Triangle Institute-sponsored Architectures for Semiconductor Integration and Packaging Conference (which I coined ASIP several years ago as I became tired of typing
out the whole phrase). This conference -- with its totally invited agenda -- gives us
a good chance to look back at what has happened during the year.
This year the preconference symposium (which has turned into course-like updates of the most recent
technological advances) consisted of myself, Erik Jan Marinissen covering test,
and Minsuk Suh of Hynix looking at the key
challenges for wide IO applications.
The keynote session consisted of Xilinx, Micron, Cadence,
GlobalFoundries, and Ericsson giving us a look at 2.5/3D progress from their
perspectives.
Keynote speakers (l to r): Vinod Kariat (Cadence), Tom Pawlowski (Micron),
Carl Engbloom (Ericsson), Dave McCann (GlobalFoundries) and Liam Madden
(Xilinx)
Micron
Micron chief technologist Tom Pawlowski discussed revolutionary
trends in memory technology and the role of 3D. He noted that "node
scaling is becoming more difficult bot for logic and memory...we are getting
close to the end of the CMOS S curve...the future will be dominated by
technologies that offer the lowest energy consumption, i.e. picojules/bit." While there are significant aspect ratio and
materials challenges for 3-D NAND (vertical transistor tech not 3D stacking),
NAND is in the process of transitioning to "3D in fab" technology since it relaxes
lithography requirements
Most practitioners feel that DRAM technology will be
replaced. Opinions range from "mid-2015" to "by end of 2019." New technologies
that may be used include:
Pawlowski indicates that the Micron HMC 3D stacking technology [see IFTLE 74,
"The Micron Memory cube Consortium" and IFTLE 95, "...Further Details on the Micron HMC..."] has in fact rearchitected memory
and equalizes signal transit time in the x, y, and z directions.
Micron offered the following memory stacking roadmap:
Longtime RTI 3D ASIP attendee Bob Patti has had a quite
eventful year at memory startup Tezzaron. Late this year they announced thepurchase of the old Sematech facility in Austin (which was owned by
failing SVTC) and the licensing of the Ziptronix direct bonding technologies
ZiBond™ and DBI®.
The Austin facility will now be known as Novati technology
which Patti referred to as "a production-style fab that can also do
development." It gives Tezzaron the control of production capacity which is something
they have yearned for, for many years.
Concerning the Ziptronix license Patti commented that "no
one technology can do it all...this gives us superior performance in die to wafer."
Patti reiterated several
times that he would be an open platform and is working with e-silicon to get
this to customers. Promising to become part of the "domestic supply chain"
Patti now has a 300mm, ISO 9000, trusted fab which can build in 65 nm CMOS
and has 6-7 interposer programs already underway.
Tezzarons
Bob Patti meets with (L) Kathy Cook and CEO Dan Donnabedian (Ziptronix) and (R)
Matt Macray (RTI organizer) and Arif Rahman (Altera)
Lester the Lightbulb
update
An engineer from the northeast who chooses not to reveal his
name (political hacks now abound in the
DOE and you certainly do not want to be called a non-believer!) sent an e-mail
earlier in the month after he stumbled onto the IFTLE Lester articles. Here is
his true story.
"...a couple
of years ago I decided to put four LED PAR38 luminaires into my kitchen ceiling
recessed cans. It took a couple of months for my wife to stop complaining about
the harshness, but we eventually settled in and even got used to the half
second startup time. Sadly, two of the luminaires failed within 18
months. Sylvania requires the original cash register receipt and UPC
symbol, and most of us aren't in the habit of saving such documentation on
light bulbs. I've since rekindled my relationship with Lester."
Hmmmmm... methinks there are a lot more stories like this out there in the naked city [1950s
US crime show humor].
By the way,
several of these Lester articles have been picked up and retweeted and
republished in the LED literature. Guess I won't be invited to give the plenary
lecture at any LED conferences, but that's OK because I'm telling you the truth
which is in short supply in some parts of the scientific community which have aligned with the politicos.
In
my one-bulb, nonscientific test which we started in Aug 2011 [see IFTLE 63, "BiddingAdieu to Lester Lightbulb" ], recall the
CFL burned out in less than a year [see IFTLE 109, "...Lester's cousin CFL dies prematurely..."]
but the LED and Lester are still both burning as of the end of 2012. Lester has now exceeded his 1000 hr expected lifetime. BRAVO LESTER ! Lester sent along this message from San Quentin where he still remains on death row but is running out of appeals.
"If you don't exterminate me I promise to be
obedient (just flip the switch and I'll come running); honest (you'll never find
me trying to bribe you to buy me or lying about my qualifications); thrifty
(hey, what else can you buy now a days for a quarter?, not my competitors, that's
for sure!); brave (I'm trying to keep my
chin up as those evil lying Washington lobbyists are accusing me of ruining our
country); and clean (when I do finally pass away simply put me in the garbage
can and I won't poison your babies if I happen to break on your floor)."
Thanks, Lester, for all you have done for all of us....
For all the latest in 3DIC and Advanced Packaging stay linked
to IFTLE.......
No comments:
Post a Comment