Evaluation Engineering - 25

every level of the process, and you must
have an integrated 3D field solver that
works at the IC and system level. This
is a tough problem that needs attention
to constraints and parasitics across the
domains.
You also need to support RF routing,
shielding, and structures (parameterized
elements) at the PCB level and stitch the
models across the platforms. You need
an integrated cross-platform solution to
accomplish this. 5G is just one driver for
this type of technology and methodology,
but certainly a very strong one. So, in a
way, technologies like this help to bring
the design disciplines together, as operating independently is no longer feasible
at this level.
RN: Do you see other industry trends
that will have an impact on the EDA
industry in the coming years?
MM: 3D printed-circuit boards constitute the next wave of prototype capability. I envision an engineer completing a
design and pushing the print button in
their CAD tool.
RN: What challenges to the EDA industry will techniques like in-mold electronics present?
MM: I think this is another example
of how intertwined electrical and mechanical are becoming. In this instance,
the case is the "board." Finding the best
way to design, verify, and build these introduces whole new processes to be addressed. From a challenges standpoint it
will mean new design methodologies and
new manufacturing ones as well.
RN: What are your thoughts on the 3D
printing of PCBs in regards to the EDA
industry?
MM: Having conductive traces on a 3D
device is becoming common. "Printing"
passive devices is feasible, but the placement and soldering of active components
is way over the horizon.
RN: How will the industry adapt to the
increasingly interdependent aspects of

electrical, mechanical, EMC, and thermal design?
MM: Through virtual prototyping that
will bring all the data and models together that can be simulated on a platform. Dassault 3D Experience is a good
example of this.
RN: Testability has always been an issue,
and EMA has partners in the design-fortest space. What can the EDA industry
do to help support design-for-test and
design-for-manufacturing in the future?
MM: A lot of this can be helped with
education and in-tool support. We need
to make design for test and design for
manufacturing part of the design flow,
not a side process. The more we can push
upstream into design, the more predicable
the design process will be and the higher
chance for ultimate success. Part of this is
educating the engineering community on
the value and need to understand these
flows up front. Just as ECAD and MCAD
are converging into mechatronics, design
and manufacturing are converging as
well. These processes are becoming very
interdependent as we continue to push
the boundaries of technology forward.
RN: The semiconductor industry has
been following Moore's law for decades,
with densities doubling every 18 to 24
months. But as one industry expert put
it, "We are running out of atoms," putting a fundamental limit on planar transistor density. Is the PCB industry facing
similar limitations, and what can the
PCB EDA industry do to support PCB
manufacturers?
MM: There will always be those that are
pushing the limits of what is possible.
From a PCB perspective, the tools are
very capable and can (in theory) let you
design in CAD what may not be effectively
manufactured using existing processes. I
think we will see more focus on manufacturing process-driven design-something
like a PDK (process design kit) for PCB.
The manufacturing process really influences so much of what is possible from a
design perspective and because of this, it

needs to be an inherent part of the design
process. The manufacturing technology
will come along just as it has with the
IC. It may not be linear scaling in terms
of feature size, but the industry will find
a way to meet the technology needs. Our
objective is to help make sure the tools
and flows are there to support the user
in a way that is in connection with all the
other downstream processes (procurement, manufacturing, PLM, etc.).
RN: Given the potential slowdown or end
to Moore's law, the electronics industry
is pursuing heterogeneous integration
(sometimes referred to as "more than
Moore"), which involves the integration
of perhaps CMOS devices and devices
fabricated in other processes. What
problem does this present for semiconductor EDA tools or for PCB EDA tools?
Do you foresee a blurring of what once
were clear lines between semiconductor
EDA and PCB EDA?
MM: I don't see this as a problem as much
as an opportunity. We are clearly hitting
some limitations in scaling combined
with the fact that the primary drivers
are not just speed anymore. Power consumption and cost are also key factors in
the equation. This leads to a new way of
thinking about what leading-edge is and
what problems it solves.
As you stated the package is becoming
more PCB-like as these kinds of hybrid
technologies evolve. I think there is a lot
for the IC community to take from the
PCB community now that these types
of heterogenous devices are gaining momentum. In general, this will mean more
integration and communication is needed
across the chip, package, and board to
drive optimal system operation.
RN: When I got my first job as an electrical engineer (admittedly a little more
than 30 years ago), PCB design was accomplished using mylar, tape, and an
X-acto knife. Obviously, the industry
has come a long way since then. Where
will we be 30 or more years from now?
MM: We will be printing PCBs on demand,
and it will happen long before 2050.
JANUARY 2020 EVALUATIONENGINEERING.COM

http://www.EVALUATIONENGINEERING.COM

Evaluation Engineering

Table of Contents for the Digital Edition of Evaluation Engineering