unless the chip making process changes significantly I see no future in the EDA.
If we look at the software development industry, most of the tools are for free (compilers, debuggers, development environemtns, etc.), and the ones which are not are very cheap. Same process will happen with the hardware development eventually.
Trying to win the price game when you have much smaller competitors in the market is not a good policy. Cadence is turning EDA into the airline business. The smaller guys can play the price game better than Cadence in the long run.
Semiconductor companies cannot produce chips without sophisticated software. Mask sets for the more aggressive processes cost a few million. Semiconductor companies will pay the bucks if the management stops being a bunch of sissies and continue to wilt to the customer pressure. Gerry Hsu held the line on the price of his verification tools in the 90's and his strategy worked. Companies paid the $$$ EVEN when they risked threats from Cadence about getting sued if they used Avanti's alleged stolen software.
Bigdog I agree but the point I am trying to make is that CDNS sales is behind the eight ball today. They have to bring in so much revenue per customer or they get fired. The fact is when they go to sell a customer CDNS tools are not that much better/worse than the other guys. Given this the only way to win is on price.
I never implied they have been innovative re-
cently. My point is that it's extremely diffi-
cult to be innovative when you spend so much
time developing "unique hooks" for virtually
every large customer. And to address another
poster's comment, this did not start with Fister, it's been going on for a very long time. Huge EDA discounts have been around since
the days of "DMV" (Daisy,Mentor and VALID) so
this isn't all that unusual. Unfortunately for
Cadence their customers are very aware that they can string their sales person along until
the end of the quarter and they can get the dis-
count they want along with whatever little
wrinkle they want to make their design flows
even more efficient. Synopsys and Mentor Graph-
ics are much more disciplined re. their sales
situations so this is less of an issue with
them. I realize the pressure comes from sales
management and even moreso from the executive
staff and that is the crux of the problem.
These clowns should be summarily fired and re-
placed with competent executives who establish
objectives that will enable Cadence to become
succesful. Then if a salesperson wants to dis-
count excessively or give away the store by
promising some "hook" for a particular customer
she/he can be dealt with by cutting their
commission by the "loaded" engineering budget
required to develop this customer's "hook".
I would suggest that this practice would elimi-
nate these unproductive giveaways post haste!
Let me try again. Sales was ordered by top management to discount and the two reasons were to boost revenue and because the discount was needed to win against the competitors. I would be interested in your view as to what products CDNS has introduced that are "compelling and differentiated" enough to replace a competitor (10 to 100 X better) or offer a completely new capability that works out of the box? In digital design, I can't think of any products CDNS has that meet these criteria. In analog, CDNS share is getting taken by small startups like BDA. This is very bad! No I won't defend how CDNS engineering gets yanked around but please don't tell me they have been innovative recently.
I see the problem differently. Cadence's sales force were compelled to sell big baskets of tools that contained
Positively differentiated products like emulators and Virtuoso where they had key positive differences WRT competitions
Undifferentiated products like digital simulation and maybe Encounter
Negatively differentiated products like synthesis, timing, extraction, test, DRC and DFM, where competitors were clearly better.
Differentiated but obsolescing products like e/SpecMan
Cadence could get good value and profitability from a few of its differentiated products if it wanted to, and be a much smaller company, but Fister chose to shoot for max short-term revenue. Hard to develop compelling products when you are busy trying to keep second and third-string products marginally competitive.
eda_animal, I submit that it is indeed the sales
organization that has created this mess along
with upper management allowing it to happen.
You blame engineering, this is the farthest thing from the root cause of the issue. History
has shown that virtually every large order com-
ing in to Cadence has a twist on it for that
customer. Each of these "twist" are unique to
that customer and is not something that all of
Cadence's customers want. Upper management,
especially in sales has almost encouraged this
practice to make the quarter's numbers! Now lets look at it from engineering's
perspective; they have to stop working on
whatever new product (or fixing bugs on an
existing product) they're responsible for to
address this "unique hook" that this salesperson
has given away to the customer to get this large
order. This is the antithesis of engineering
productivity. You talk about "highly differ-
entiated and compelling products, you'd be
amazed at the "compelling and differentiated
products" engineering can turn out if they don't
have to spend an inordinate amount of time
cleaning up a bunch of salesmens/womens "give
aways". Obviously sales management and executive staff are the one's who allowed this
to proliferate over the years, but "yes, Virginia thre is a Santa Claus" and to eliminate
this problem you need to let engineering do
what they get paid to do and that is to
DEVELOP NEW INNOVATIVE (EVEN COMPELLING) PRODUCTS!!!!!!!!!!!!!!
The other part of the solution is to remove this
BOD who have allowed this practice to continue
I agree with much of what you say but please don't blame CDNS' sales. The reason they had to discount so heavily was that CDNS engineering did not build highly differentiated products. I’ve worked at other EDA companies with very compelling products that customers would pay top dollar for. Cadence just does not build products like this.
semicon20ye and lesterfudd have hit the nail
right on the head. EDA utilized the premise, right from the start (back in the early
80's), if they could provide something that someone could throw at the wall and it stuck,
they would sell it. Unfortunately for CDNS the
customers started demanding "working products"
(imagine that) and MENT and SNPS realized this
and "grew up". CDNS, however, was and is obsess-
ed with the SV mentality and continued to ignore
their customers while they all lined their pock-
ets by promising the world to their customers
and giving away the store to get a sale. Eng-
ineering was always frustrated with this
approach because they spent an inordinate amount
of time developing unique hooks for specific
customers to get some order that a salesman
wanted to fill his quota. Therefore, a lot of
engineering time went to creating frivolities
rather than state of the art products that all
customers could use. As I say their competitors
learned this lesson along time ago, but CDNS
(especially sales) continued to act like a bunch
of whores at the expense of the rest of CDNS and
their customers by continuing the same approach
and they have finally paid the price for this
rediculous strategy. The last guy who half way
understood what the customer wanted was Costello
and the BOD took care of him. They then com-
menced to bring in a string of bozos who would
follow the tried and true CDNS approach of lining their pockets with cheap option grants. So, I agree with the fact that CDNS needs to
bring in a new CEO who doesn't have the SV
mentality (Mentor brought in Rhines from Texas
Instruments years ago). However they need some-
one who is first and foremost going to address
their customers needs and flatten this organization such that they get back to being
a productive entity where they are all pulling
in the direction of developing and providing
world class products for their customers.
This will never happen with the current BOD.
These clowns have raped this company for more
than 20 years. They have allowed the company
to operate in this catastrophic manner while
they have become rich off cheap option grants.
Therefore, for CDNS to be successful, you not
only need to clean house at the top, reorganize
each department such that they get down to "fighting weight", but first and foremost get
rid of this incompetent board of directors.
The class action lawsuits are interesting be-
cause they only target the top employees. Why
isn't the BOD included in this action as well?
I realize they are protected, but there is
nothing wrong with identifying them as the
major source of the problem. Hence, lets see
a completely reorganized CDNS structure, with
an insightful (one who is serious about solving
their customers problems) CEO and a new and
competent board of directors. This company
used to be the star of EDA and can be once
again with the right leadership!
Another dumbass comment. Software tools are not free - the most widely used s/w development environment is probably Microsoft Visual Studio which costs several hundred dollars. Even "free" tools such as Java from Sun are really subsidized by the sale of Sun hardware. EDA is a multi-billion $ industry - you can question it's growth, but there are companies doing quite well, and others not so well. Take a look at the SNPS balance sheet.