Here is a point of view of a person who has
worked in this industry for a while:
Synopsys has had a stranglehold on ASIC synthesis in the
past, the market is changing as follows:
- in the
FPGA synthesis area, a startup is cleaning Synopsys'
clock, and FPGA synthesis is supposed to grow as today's
FPGA rival yesterdays ASIC in size
- this same
startup will launch a ASIC synthesis product which will
be superior to Synopsis, 10X faster and LOWER in
price. What will happen is that a) customers will knock
down Synopsis ASP b) buy fewer Synopsis licenses c)
Synopsis will lose market share in ASIC synthesis. This
will not happen overnite, but in 2 years from now
Synopsis could easily lose 25% market share and in 4 years
2) Cadence offering Synthesis at low price - people
do not buy CDN synthesis but use it to knock down
Synopsis ASP. (av selling price)
3) Synopsis has
been great in past, but it may have peaked, even in
terms of technology
-in simulation, NC Verilog has
caught up with VCS and is even faster in some cases, so
there is not big technological advantage to Synopsis _
I do not see it gaining market share
Synthesis I see Synopis losing market share in future
because of above discussion, but also fact
Synthesis without link to P&R is not a viable solution for
tomorrrows ASICs. In this market there are multiple players
besides CDN and SNPS and it is not clear who the winner
is. One cannot assume just because SNPS has the pure
Syntehsis market, it will prevail in this area. Over a
dozen startups are betting that they can take away
dollars from SNPS.
- In P&R Synopsis has been a
non-entity. With EPIC they got some good tools, but many of
those people are leaving to seek bigger bang elsewhere
which typically occurs after an acquisition
for CDN, things are not as bad as some people seem to
think on this board. They have equal product in
simulation, and they have very strong product line in PCD
(ORCAD) and in emulation (Quickturn) which are both doing
well -where SYnopsis has no presence. The ASIC P&R
area they have lost their stranglehold of past years,
but CDN mgmt realizes this and getting back to good
shape for the company does not require regaining market
leadership there - the PCB/Emulation sales far exceed their
P&R revenue. Also CDN is diversified into methodology
services and Design services
I would say, in
in near term (3-6 mos) CDN is less risky than SNPS.
CDN may not gain more EDA dollars market share, but
will not lose it either.
in long term, SNPS
will not be the dominant EDA player it has been
because its Synthesis leadership positon will erode. It
may still comeback because it is oversold, but do not
expect it to be the guerilla it was (CDN was also a
guerilla one time ago). SNPS will lose EDA dollar market
share. SO beware of looking at Synopsis just based on
historical financial/market data.
EDA is cold and
networking is hot - lot of people have left EDA for .coms
and NETW companies. The .coms have fallen and the PEs
NETW may not last 5 years from now. The EDA tool
licensing revenue model is suspect and these companies may
never have spectacular growth and PE but this is
absolutely an essential industry enabling the CSCO, EXTR etc
to design their stuff.
Many engineers are not
very stock market smart and risk taking and especially
the junior ones will not feel excited to join a
company because it is at 52 week low - they will accept a
job across the street where the stock is at 53 week
high. Let us face it, in spite of all the msg boards
etc, most people in stock market buy high and sell low
(individuals). Over last few years with teh run up people have
been lucky to buy high and sell higher, but most
people are not suited for risky turnaround plays which
is what buying SNPS stock or joining SNPS
If edaguy is counting on Synplicity to be a
challenger to Synopsys, he's should look at Synplicity's
recent S1 filing. Looks like they will only be able to
compete by losing money faster than they do today. What's
facinatign is that they are positioning themselves as a
"software supplier to the internet industry", rather than
as an EDA company. I bet their IPO drops through the
floor once investors figure out what's real.
Berndt Braune was formerly one of the @$$wipes
GMs from Mentor who thought that selling the Meta
emulator that infringes on QuickTurn patents might be
defensible, good idea. Well Mentor lost with the ITC, lost in
court and Bernde lost his job (funny how that works).
His claim about former SNPS executives going to
Get2Chip shows the same kind of truth-stretching that got
him trouble with Mentor/QT - What Berndt really means
is that some former SNPS guys went ot Meropa/G2C and
were made executives.
It is strange to see a detailed forecast of
Synopsys (written in previous msg Synopsis) from a person
who isn't even able to spell the name. This board
gets more and more the playground of people trying to
manipulate others for personal profit and not to exchange
Well said. As a former Synopsoid, I feel the
company will bounce back to the high 50's low 60's just
based on its cash in the bank $800 million, and
revenues ($1 billion). At its current valuation, it would
behoove some deep pocket finianceer to buy SNPS and sell
it off piecemeal and raid its piggy bank.
edaguy's post has some meat in it. However the
issues have been around for sometime and the company has
been doing fine. Synplicity has been suggesting about
releasing their ASIC synthesis tool but if Ambit was an
example, they have no chance of getting into this market;
the buzz right now is Physical Synthesis. Ambit
needed CDN's sales team to make a small dent;
Synplicity's ASIC tool will have a tough time doing it alone.
I do not think any business-savy CEO will go in and
release it; too much investment for too little rewards.
Regarding FPGA synthesis, true Synplicity's tool is
possibly preferred ; however for most engineers the OEM
Synopsys FPGA is good enough (and virtually free). The
FPGA market is growing faster than ASIC but there is
are performance limits to where it can go, simply
because of technology constraints like power which limit
the speed at which your best FPGA can run.
Regarding cheaper synthesis alternatives, Synopsys is
responding by offering term bases, subscription based and
even the Internet model to offer a competitively
priced alternative while allowing users to stick with
the leader. True they will lost an account here or
there, but the real battle is in the large accounts
which are sticking with SNPS.
NC-Verilog/VCS etc: Each tool's performance moves up in step
function with each new release. However people do not
migrate en-masse from one tool to another just because
one tool performed better in some benchmarks. There
is no compelling reason for people to move away from
VCS. However, over the past couple of years when the
verification crunch hit, a lot of people have invested in VCS.
These customers are sticking with VCS and as the size
and number of chips grow, will be buying more.
Starting from the CEO, CDN is producing a lot of FUD;
however they are yet to produce a credible tapeout story
for the SPR product or get customers to say that we
have switched from xxx to NC-Verilog. CDN would be
willing to replace the xxx licenes for peanuts to get an
endorsement like that. Synopsys has not done a good job in
responding to the questions sorrounding the company. Part of
the reason may be the basic company culture and
ethics; people do not want to make statements which they
cannot back-up. This is very visible in the press
releases about the products: PKS announced nine months
ago, yet to see a tapeout till now, Physical Compiler
announced with two high end graphic customers talking about
their experience and tape-outs.
From your analysis, I'd say if you ever worked in
EDA it was about 1-2 years ago.
is old news. They were supposed to have a hot new
ASIC synthesis tool over a year ago. Nothing happened.
If they did have something that was 10X faster (with
equal or better results which would be a requirement),
why would should they sell it cheaper? Cheaper price
is how you sell an inferior product
2. Yes, customers can get the job done with either
VCS or NC. However, with designers in a verification
crunch, SNPS sells VCS in bulk and it is still #1 in new
license sales (and growing market share).
Yeesh, you really missed the boat on this one. Haven't
you been reading for the last year? Synopsys is
offering Physical Compiler which ties synthesis and
placement together. Over half a dozen customers have taped
out production designs done entirely with PC (unlike
CDN who can only claim one partial tape out).
The industry is re-tooling. IF SNPS can win the
physical synthesis war (they have a 9 month lead so far),
this should have a big impact to their bottom line
around 12 months out. Your information is flawed and out
>2. Yes, customers can get the job done with
either VCS or NC.
However, with designers in a
verification crunch, SNPS sells VCS in bulk and it is still #1
in new license sales (and growing
I'd check your numbers again if I were you, because
the EDA analysts know
that NC-Sim, NC-Verilog, and
NC-VHDL have been selling in bulk for some time,
customers are replacing wholesale their VCS licenses for
even more NC licenses.
CDN share is growing faster
than the market in this area ==>> read: SNPS is
going down and can't stem the tide.
positive territory. As for the posters on
technical merit, please keep up the debate. As much as I
hate to hear bad things (and even bad things that are
false) I get a kick out of replies with corrections to
falsehoods and new info. Thanks guys (or gals).
Synplicity, the not-so-young startup you
mention by name in the FPGA synthesis
area (and the
one whose tools we use),
has a big problem though:
not liquid. And Synopsys has a pretty
good hold on
the FPGA synthesis distribution
the major FPGA vendors.
So actually the most
path for Synplicity is to get bought
If you were Synplicity, who would you
to get bought by? They're not going to be big
to IPO! Avant? Yech.
Only thing standing in
the way on both sides is pride.
But engineers on
both sides are smart people and
would respect each