Recently laid off, I visited my colleagues at Cadence building 10 yesterday. I was very impressed with the new building. Visit it if you get a chance. I hope with this new environment, Cadence R&D will be inspired to bring this company back to its former performance. It's worlds better than the River Oaks campus, where I worked for 15+ years.
I was talking about the future of EDA, not its glorious past. You yourself admit that technology shrinkage will come to an end. And that will as well bring an end to eda as we know it today. It is always painful to see pass away something we love, but that is the future of EDA. Open your eyes and see it. The sun is setting for EDA, not rising.
Let me pick up on the edaisdead argument....
>some more fancy tool options and catching up with ever shrinking >CMOS process but to me that does not comprise innovation. >Innovation would be something that completely changes the rules >of game
Since 0.18u it has taken more than a few fancy options to make efficient use of silicon. Fundamental algorithmic changes, and merging of problems / algorithms was required.
In my opinion geometries R&D consider are/will tail off. We are at a fundamental shift, where multi-core will take up where shrinking geometries can't economically deliver.
The driving force has never been shrinking geometries. It's been processing power per $.....
Dog eats dog is nothing new, look through the changes in the landscape in the mid eighties.
EDA IS an integral part of SOC design if you look at the
software used for design, design verification, physical
implementation (floor-planning, transistor placement and
connection, or place&route) physical verification, design
for manufacturing, manufacturing s/w, board design, board
manufacturing, and the list goes on. Big Semicon used
to develop and support all this s/w themselves, but the
cost of development and maintenance was too high for them
when you put it up against other costs like engineering,
manufacturing, etc. What these companies figured-out is
that getting rid of EDA tool development and getting
rid of their manufacturing facitlities, etc. and just
concentrating on chip and hardware design and letting
EDA companies develop the tools and fabrication facilities
(like TSMC, UMC, Chartered)do the manufacturing, they
can specialize and cut huge costs, instead of trying to
take-on everything. They partner with EDA companies to
either develop new tools based on their needs coupled
with the fabs, or to enhance the tools that are already
there. Either way, it takes innovation to do both in
keeping-up with shrinking technology nodes, new processes
and needed design styles (wireless, analog-mixed-signal,
digital, memories, etc.).
What is killing Cadence (or possibly has already killed
Cadence regardless of the "good stuff happening inside")
is greed, pure and simple. Their burn-rate is just too
high ... executives spread all over the country who fly
to headquarters weekly, stay in corporate appartments,
rent cars, fly 1st class, etc. and even after they fired
the biggest monkies on their back, they till have this
problem. Ask a CDNS employee why they're still there in
light of the last year's events, and it's "because I
make too damn much money". Ask their competition, and
it's "the company is making technological advances ..
it's an exciting time".
"EDA is dead" is like saying the manufacturing sites in
the world are dead, or design-houses are dead. They all
keep themselves going ...
If we look at EDA as an integral part of SOC design, then there is no doubt that we cant tape out chips without these s/w tools from cadence and competitors. But we consider EDA as engineering discipline and business, I believe it is dead. To call something alive, it has to show growth in buisness and some innovations to keep the show running.
With 4 billion dollars market and ever increasing competiton between the top players, it is a dog-eats-dog kind of scenario. So one's gain is always gonna be other player's loss. (And for cadence, the competitors are fast catching up with its strongest bastions Virtuoso et. al.)
Coming on to engineering (again the digital design part), there is little scope I see in further innovation and one proof of that is fast drying up pool of startups in digital design and implementation arena. Alright, we may see some more fancy tool options and catching up with ever shrinking CMOS process but to me that does not comprise innovation. Innovation would be something that completely changes the rules of game. Tool modification that is happening in EDA industry is more a question of survival, otherwise other "dogs" are waiting to eat what you have. So EDA is DEAD :)
A new building doesn't indicate anything other than Fister spent a lot for a new building. The company needs solid leadership and an insightful technical direction.
Where is the innovation? I haven't seen anything new come out of Cadence for years.
You can't have been looking very hard!
Look a bit harder at the products or monitor the innovation awards.
Cadence is not a startup. Your view does a disservice to the many smart engineers
that work in Cadence, Synopsis etc
If they aren't innovating what are they all doing??? ... - and no the answer is not full time bug fixing.
Re: stock, With a market cap difference of 3x, we are way off the normal swing of the pendulum. Just watch and you'll see, it will return.
well you must be getting rehired
new buildings not worth much if there's no talented individuals in them
too many managers spoil the broth mate
hire back the tech wizards they fired
once the customers start to use the tools they'll start to find out the true colours
no surprise they'll switch to purple instead of black with the red line above the "a" if ya know what I mean
I'm not (yet) getting rehired, but I would love to. And I think there are many talented people still at Cadence. And I know many of them. They have a tough road to hoe, but I would love to help them with that. Like I said, in that bright new building, it's like a rebirth for R&D.