La Monarca Vol 4 FINAL - Page 4

M i g u e l Á n g e l L a s t r a s M o n ta ñ o
Continued from page 1
Photo courtesy of Miguel Angel Lastras Montaño
Surprisingly, the memory in
our computers is essentially old
technology. Originally developed in
the 1970s, complimentary metaloxide semiconductor technology
(CMOS) is reaching its physical
limits. In Miguel’s eyes, modern
personal computers are like digital
combustion engines are optimized to
be incredibly efficient and powerful.
However, the technology is essentially
the same as what propelled the Ford
Model T over 100 years ago. Like the
internal combustion engine, CMOS
technology has been pushed to its
theoretical limits of performance. As
Miguel explains, “With the current
way the main memory performs in a
computer, we’re reaching a wall.”
The problem, he says, is that
traditional chips are essentially
two-dimensional structures: “The
industry is very comfortable with
CMOS chips, very resistant to
change. The flat chip is entrenched,
and isn’t going anywhere.” But
expanding into a third dimension
could allow orders of magnitude
more information to move efficiently
for a given area, like replacing
stoplight interchanges with flyovers
and overpasses.
found a way to make amends with
the flat-chip folks by building 3D
chips right on top of the old ones
and finding a way for the chips to
communicate seamlessly with each
other. Adding to this complexity is the
new and future need for his creations
to be increasingly energy-efficient
and miniscule due to exploding
demand for mobile devices.
Yet it’s not all theory. For his
dissertation, Miguel successfully
designed a 3D memory chip
that colleagues at UC Santa
Barbara, University of Michigan
and University of Massachusetts
fabricated, for which a working
prototype now exists. While Miguel
can no longer think in 2D, he also has
Like many successful academics,
Miguel’s path to his current field
of research was circuitous. As an
undergraduate at the Universidad
Autónoma de San Luis Potosí
(UASLP) he studied engineering
physics. He took classes in quantum
mechanics and physics but then
became interested in parallel
computing after securing an
internship at IBM. Believing he had
found his passion, he completed a
master’s degree at the UASLP and
later applied to UCSB to perform the
same kind of work he did at IBM.
Finding the right advisor proved
challenging, and ultimately he
changed course to study memory
chips and hardware with Professor
Tim Cheng. In retrospect he says this
change of paths was very fortunate
and placed him at the forefront of an
exciting new area in computing. With
exponentially increasing demand for
faster and smaller computers, Miguel
is confident that if you build a better
CMOS chip, the world will beat a
path to your door. •

Powered by

Full screen Click to read
Paperturn flip book viewer
Download as PDF
Shopping cart
Full screen
Exit full screen