( ESNUG 438 Item 18 ) ------------------------------------------- [01/18/05]
Subject: 18 EDA Employee Reactions to the Cadence-Verisity Merger
> Cadence announced it's buying Verisity for $285 million in cash.
>
> As an EDA vendor, do you think this is good news or bad news? Why?
Paying $285 million for a proprietary testbench technology that is
going the way of the dodo because the world is going to System Verilog
is just stupid. The hardware acceleration and code coverage tools
don't begin to make it worth this price. Congratulations to Cadence
for visiting the ICU to buy another company.
- [ An EDA Employee ]
From Cadence employee point of view, this looks good but it depends
what our management will do in the close future with Verisity
technologies. I already work on Verisity software, especially Specman,
linked to Cadence Palladium (both work excellent separately) but the
interface was not so optimized, so I believe we could merge our
technologies in order to create something really excellent.
- [ An EDA Employee ]
What the hey? Duplication of tools for CDN = more headaches for CDN.
So is it is access to e seats? Don't think so when proprietary
verification languages are dropping like flies. So is it access to
the Axis seats? Hmmm, I think CDN already billed itself as "the
leader in verification". Was there some talent at VRST they needed
to buy? Dunno.
Basically, I'm perplexed. Maybe it was someone's MBO to "acquire one
company per quarter" and it was either this or lose a merit increase.
Clearly VRST needed to have this happen 'cuz they weren't one of the
Big Four, but what held a gun to CDN head and said "do this"? That
I can't figure out.
Makes no sense to me.
- [ An EDA Employee ]
E should die and this prolongs its life unnecessarily.
- [ An EDA Employee ]
The acquisition will cause Cadence trouble because they will have
to decide what to do with E -- keep it or scrap it. I vote for
scrapping of course. The EDA world just doesn't need another language
war. EDA spends too much resources on languages -- Verilog, VHDL,
SystemC, System Verilog, OVA, PSL,.... It is a waste of resources.
And language is pretty much religious ground. Lets improve algorithms
and tools. There, I said it. I feel much relieved.
- [ An EDA Employee ]
As an EDA vendor it doesnt affect us cause we don't compete in that
area but I am sure other verification EDA vendors are happy because
CDN has acquired Verisity, which means traditionally the product will
lose a couple of quarters easily in assimilation/reorg etc. It
happened in case of Ambit, HDL systems, Get2Chip, etc. Good thing
is the people in Verisity will be happy $$ wise!!
- [ An EDA Employee ]
I can not understand why they would pay so much for VRST. Working in
the field I see customers moving away from 'e' and Verisity and towards
System Verilog or SystemC -- both of which are free. Cadence certainly
didn't need the extra head-count.
The only conclusion I can come to is that they are going to try to
establish 'e' as a longer term solution for customers as a hedge
against System Verilog. After all SNPS still has Vera.
It is too soon to say if this is good or bad news for EDA vendors, but
it does serve to confuse customers in the short term.
- [ An EDA Employee ]
For vendors and Wall Street: Good. Consolidation is good for such
a small industry. Expect more to come.
For Verisity Customers: Good ... maybe. Depends on how Cadence R&D
manages the tradeoff between legacy support and System Verilog support.
- [ An EDA Employee ]
As an EDA Vendor looking for funding, this has to be good news. It
shows return potential and that EDA isn't a stagnant market. I haven't
seen the reaction from the VC's (or potential acquirers) we're engaged
with yet, but I have to imagine it will be positive.
- [ An EDA Employee ]
I am surprised to see Cadence announce the acquisition of Verisity for
that high amount. I am curious to see if the deal goes through like
"Avanti" or fails like "MoSys".
Overall, it is good for the small startups that the BIG companies are
shopping again and willing to pay high amount for acquisitions. A
NICE way to start the year for EDA startups.
- [ An EDA Employee ]
I think it is good news as there is now a consolidated 3-way battle,
there is enough competition to keep an efficient market. The open
question is how well Cadence can integrate the VRST R&D team, which
is primarily based in Isreal. My prediction is that this will not
go well and there will be a major culture clash.
Specman will stay stagnant and either Mentor or Cadence will win the
simulation war. Cadence will be relegated to the hardware-only space.
Industry needs to have a single consolidated design and verification
language and it is System Verilog. CDN now has even more languages
(SystemC, Verilog, VhDL, PSL, e/temporalE, System Verilog). It's a
very crowded house. This will make it hard for Cadence to achieve
alignment and focus both internally and in their customer base. My
prediction is that they will keep all doors open like Mentor and the
industry will not align on any one language.
- [ An EDA Employee ]
You have asked the $285 M question. The news release sounds rather
positive with "join the Cadence(R) executive management team" and
responsibility for "an integral role in setting Cadence's verification
technology direction", but these are not top positions.
My bet is that Cadence will have both options on the table for some
time while they let the internal and new (Verisity) efforts compete.
- [ An EDA Employee ]
EDA growth and revenue is virtually stagnant. This acquisition
provides Cadence with two things: Marketshare growth and Cash.
Verisity, based on their Q304 report, had ~$54M in cash on hand.
Besides market share and cash, Cadence finally gets a complete
verification environment. However, they are stuck with duplicate
simulation and hardware emulation technology plus approximately
300 Verisity employees who, for the most part, duplicate existing
Cadence personnel.
Synopsys has changed the overall verification rules by bundling VCS
and Vera as one complete offering. Cadence can now offer similar
bundling of NC-SIM and Specman, but at what price?
Overall this deal makes sense only if Cadence is looking for immediate
gratification, i.e. marketshare and cash. Over the long term, IMHO,
the issues around sustained revenue and increased marketshare are far
outweighed by a shrinking hardware market, downward spiraling ASP's,
and merger integration issues. Over time I suspect Cadence
shareholders will be very unhappy at just how much this impacts
their share price.
- [ An EDA Employee ]
Another EDA industry turkey buying an almost dead company to make it
appear they are not going out of business. And the business CDN is
buying is based on "e", a language that is all but dead in the face
of System Verilog, SVA and the System Verilog Test Bench.
Why are they doing this? Because the heads of Cadence were brain
dead a long time ago. Why don't they all just go to work for the
government?
- [ An EDA Employee ]
Verisity is a very non-conformist company. Several of the employees
are looking for employment in startups.
In general the EDA market is shifting again towards consolidation.
Remember in the late 80s, when Cadence bought Valid, and mid 90s when
everyone was buying simulators: Speedsim, VCS, MTI, etc? Verisity
did not stand a chance with its current pricing scheme, and Vera
was gaining on them
All in all, it seems like the right thing to keep Verisity alive
(at some form).
Another question is what will happen to:
Verisity sales force
Axis products (seems like each have a Cadence equivalent)
SureCov products (do these exist?)
- [ An EDA Employee ]
I see this acquisition as positive for verification users because
instead of a bunch of small initiatives and products confusing the
market, this will probably polarize the vendors and users around
2 main approaches (Vera and e). With many of Vera features in
System Verilog and 'e' having been proposed as a donation to IEEE,
there is little scope (and need) for other approaches to gather
sufficient momentum for success. This is good news for users as
weaker approaches will disappear and reduce the clutter one must
wade through in selecting a verification environment.
Question is - how does Cadence handle overlap of products such as
Quickturn and Axis emulators, Specman vs (testbench tools) vs SystemC
support, and embedding Specman/'e' features into NC-SIM/verification
products. Vera as a product became almost irrelevant once VCS
embedded 'direct access', and System Verilog included as its feature
set. The usage lives on, but in a morphed form.
I give it thumbs up for users, but neutral for Cadence business. Of
course, double thumbs up for $12/share cash offer to those hard-working
Verisity folks!
- [ An EDA Employee ]
I do believe that Cadence makes sense. Mentor would have made even
more sense, but I am not surprised given the size of the deal that
Mentor didn't get Verisity in the end.
- [ An EDA Employee ]
As an EDA vendor, I think this will kill some of the rapid progress we
have seen over the last years in functional verification. Verisity's
presence did put a positive competitive energy on all involved. Being
"sucked-in" will kill most of that.
- [ An EDA Employee ]
NEXT->Item
|
|
