Get email delivery of the Cadence blog featured here
Rather than wandering too far off-track with this one, what
celebrity is more well-known for successfully reinventing themselves than
Madonna? And it's probably less about reinvention than it is about adapting to
a changing marketplace. How many other 1980's pop stars can still sell out
We also have some great "reinvention" examples in our
industry - IBM transformed from mainframes to PC's to software and services.
Apple transformed from PC's to consumer electronics, and it could be argued that
what they're really doing is creating new ecosystems. EMC has transformed from
primarily hardware-driven to primarily software-driven, while still servicing
the same need.
It's how you define yourself - what you do, versus how you
On the other side is the railroad companies of the 1800's -
instead of focusing on the "what", which is transportation - they focused on
the "how", which was trains. They missed the transportation's technology shift
to automobiles and airplanes. Digital and Data General were the top
minicomputer companies, but as the computing needs shifted, they did not. We've
seen the major "record labels" define their business as selling CDs and thus
resist the consumer-demanded shift to MP3s and streaming audio. Really they
should define their business as the experience of music. Similarly while
Madonna defined herself as an entertainer and thus adapted over time (even
acting in movies), pop stars such as A Flock of Seagulls and Tiffany did not
endure. "Pop" by definition is something that is always moving to the new
What does this have to do with chip design? How do you
categorize what you do? Do you consider yourself an "RTL designer" or an RTL
synthesis expert? Those are examples of "how", just as simulation is to
verification. What happens when SystemC-based Transaction-Level Modeling (TLM)
Do you try to sabotage the migration like the record labels?
It sounds silly, right? Then why do we have so many physical design customers with
no synthesis experience using RTL Compiler to improve netlists produced by "DC
So it seems we need to define ourselves around the "what" we
are doing, rather than the "how" we get it done. Even just focusing on logic
design, hardware design, or chip design, may be too narrow. Today's electronics
products require the hardware and the software be designed together. Perhaps if
we want to age like Madonna, we need to have titles more like "cell phone
designer" or "network storage designer".
Music artists who mastered Reinvention?
Prince. And Cher (who has had a #1 hit in each of the last 4 decades)
Anyhow, I digress...
I see your points about considering the "what" and not the specifics of "how" its accomplished. I think when we can take each iteration of of a "what" and learn ways to improve or raise the level of abstraction at which we deal with problems, that will open new doorways.
We'll be able to predictably design more with less effort. And leaps of abstraction come along in 5-10 year cycles. It's time for a great leap forward. (Billy Bragg reference).