Drama in EDA

Whistling into the wind

by Peggy Aycinena

Before I sat down to write up the EDAC CEO Forecast Panel that took place last Thursday evening in Palo Alto, I crossed paths with a guy close to the EDA industry who gave me a severe talking to.

He told me that the problem with EDA is that the only thing the EDA press ever does is repeat rumor and gossip – that they don't do what they should be doing, which is to focus on the technology and the customers.

He told me that all the EDA press ever does is look at Cooley's site and repeat rumors that they see there, that Cooley's not really doing analysis, he's just providing a billboard for people to post their opinions. He told me the only person doing what should be done, is Jay Vleeschhouwer, because Jay's really talking to the EDA customers and finding out what's working and what's not.

I said, "But Jay Vleeschhouwer's not with the press."

"Yeah," he said "But he's doing what the press should be doing. He's talking to the vendors and then he's talking to the customers."

"But," I said. "The vendors don't really want the press to talk to their customers. They want to filter the entire message and make sure it's got the right spin, the right 'first-time success' story in place before the press hears about it. And besides, that's the only kind of press the vendors are willing to support through their advertising."

"Yeah – well, the people in the press should have their own internal contacts at the customers. They should be able to call up their contacts and find out the real story. Like Jay does. The only other person who does it right is Gary Smith."

"But," I said. "Gary Smith's not in the press either."

"It doesn't matter," he told me. "He's doing it the right way."

I asked him: "So let me understand – the only thing wrong with the EDA industry is the press?"

"Yeah," he said. "The dynamics in our industry are no different than in any other industry. It's just that the only thing the press does here is look around inside the industry, and at the rumors. But that's not where the story is. The story is outside the industry – where the technology from this industry is being used. And that's why Wall Street doesn't respect us. What the press should be doing is what Jay's doing."

****************************

Okay – having been fairly warned, here's what I saw Jay doing last Thursday night.

****************************

Jay Vleeschhouwer is an analyst from Merrill Lynch who tracks the EDA industry. He was on hand last week to moderate the annual EDAC CEO Forecast Panel. The event is held each year somewhere in Silicon Valley and is described as the place where "the CEOs of the leading EDA companies can share their impressions of the business outlook for the coming year."

This year's leading EDA company CEOs included:

Aart de Geus, Synopsys
Jim Douglas, Reshape
Mike Fister, Cadence
Wally Rhines, Mentor Graphics
Chris Rowen, Tensilica
Sanjay Srivastava, Denali Software

Although Rajeev Madhavan heads up the fourth largest EDA company, he was not there. And although Tensilica just joined EDAC this past week, Chris Rowen was there. The presence of the other people on the panel was not a surprise.

Once underway – after cocktails from 6 to 7 – the panel lasted from 7 to 8:30 and included opening remarks from Jay, a chance for each CEO to address the audience of 150 people, and a chance for the CEOs to then react to questions from Jay. No questions were taken from the floor and most of the CEOs had PowerPoint presentations to back up their opening comments. The event was taped, so perhaps you'll want to watch the recorded proceedings after they're posted on-line to get the true picture of the event.

Jay Vleeschhouwer was introduced by Hal Bennett from Novell. Novell was a sponsor of the event. The other sponsor was Hewlett Packard, which was probably why the event was held in the HP corporate headquarters off Page Mill Road in Palo Alto. Jay made these points:

* The industry should be examined from the point of view of incentives.
* The technology is complicated and hard to understand for those outside the industry.
* There used to be excitement and fascination on the Street about the industry.
* The excitement and fascination needs to be re-instilled.
* The technology and tools have changed.
* The tools of the trade for analysts have changed as well.
* Novell has impacted that technology.
* Companies that don't meet expectations, don't succeed, and he sited Daisy from 20 years ago as an example.
* Analysts consider the following in evaluating a company:
… Structure
… Attributes
… Drivers
… Business models
… What changes and what can change in the technology?
* The EDA industry is highly concentrated.
* The large vendors dominate.
* Compared to other industries, EDA exhibits a unique concentration of a few large vendors.
* The number of users of EDA software is relatively small.
* There isn't a mass market for EDA software.
* The EDA sector is orders of magnitude smaller than other software sectors.
* The EDA industry ought to be a vital industry.
* One would think the profitability would be greater.
* With $3.8 billion in gross annual revenues, there's a disconnect to what actually happens.
* Or else, why wouldn't there more profitability in EDA.
* Just look at the 6-figure deals and bookings up front that characterize the industry.
* You need to look at the technology needs, the capacity, and the market economics within EDA.
* There are many diverse categories of tools in the EDA industry.
* Dataquest counts 5 dozen categories.
* There's a lot of fragmentation within each category, so each company is a leader in its niche.
* Everyone knows everyone in this industry.
* Can everyone be held accountable for the current state of affairs?

Mike Fister from Cadence was next. He said:

Sanjay Srivastava from Denali was next. He said:

* It's the back end.
* Customers needs software validation.
* Customers needs DFM.
* He showed an SoC Enablement Report Card that listed successes.
* The list included PCI Express.
* What is needed are effective standards.
* Potential areas for growth in the industry include:
… IP
… ESL
… DFM
… Platform enablement
* The unsolved challenges include:
… Software product gaps
… Interface enablemenet
… Channels for designing with IP

Jim Douglas from Reshape went next. He said:

* What drives consumers are:
… Gas prices
… Employment
… The increasing costs of food stuffs
* How does all of that affect the structure of the EDA industry?
* The Adam Kablanian Theorem was sited to describe the EDA industry.
* Instead of there being 8300 airports in the U.S., there would only be 4.
* At 65 nanometers, it's one thing to build it, another thing to support the applications.
* The question is when is there ROI?
* It takes 2 years to re-tool.
* The cycle tracks Moore's Law.
* Semiconductors leads to ROI.
* The big question is – EDA is tied to semiconductors, so therefore also to the semi ROI?
* R&D for big EDA vendors at an all-time high.
* There's a ton of pressure on the big vendors.
* There are great opportunities at 90 nanometers, and ESL tools across the board.
* And there are new business models at 65 nanometers.
* The industry will be relatively flat this year.
* From –5% to + 15% growth.
* For electronics, this will be a great transition year.

Wally Rhines from Mentor Graphics went next. He had a lot of graphs and he said:

* There are unique things about the EDA industry.
* It's unlike other software industries.
* We've had a gradual transition from perpetual licenses to 3-year licenses.
* It's had negative impact and half the TAM has been destroyed.
* The benefits:
… Continued revenue stream
… Compelling renewals
… Predictability of growth
* Today the business is renewing 3-year lease with maintenance.
* Increased tool usage is reflected in a renewed commitment.
* There's a fixed renewal date at the end of 3 years.
* Oracle and PeopleSoft use a 3-year lease + maintenance through the life of the contract.
* Cash flow has been dependable over the last 3 years.
* The company has had an astounding repeat in cash flow.
* There has been predictability with a 3-year renewal cycle.
* EDA are in an upward trend.
* The EDA industry peaked in 2001 and declined afterwards.
* But R&D spending is up.
* R&D spending will increase even more.
* The 2002 licenses will renew in 2005, and then usage will go up.
* The forecast always waits for the quarterly EDAC MSS numbers.
* Fourth quarter numbers indicate an approximately 3.3% annualized growth for 2005.
* But 2005 will grow.
* 2006 will be even better.
* In 2006, the number will be 8.5%.

Aart de Geus from Synopsys went next. He said:

* The industry statistics require an air-sickness bag.
* There are dramatic changes underway.
* On a global basis, there are technology changes.
* Moore's Law is not changing.
* The ROI has changed.
* Technology and economics start to touch each other.
* That is techonomics.
* Compare today with the Industrial Age in the late 1800's.
* The number of people now living in free markets is increasing.
* However, their spending levels are much lower.
* Therefore consumer products are even more price sensitive.
* These past few years have not been a downturn.
* They have represented a secular shift in the growth rate
* The numbers in EDA will be 5% to 10%.
* The consumers are very cost conscious.
* EDA must therefore reduce costs.
* The industry has to look at the whole economy.
* And also look at the entire flow.
* Wally's graphs were astounding.
* 46% growth in 2001.
* 36% growth in 2002.
* Right now, the industry is still flat.
* But there are $3 billion in backlogs for Cadence and Synopsys.
* The ability to invent in EDA is amazing.
* Designers want the industry to stop showing the X, Y, Z tools.
* Designers want the industry to show product solutions.
* Productivity = Revenue/Expenses.
* EDA can help by cutting tool costs and cutting customers' costs.
* EDA is small, but has a huge impact.
* This year the growth will be 5% in excess of the semiconductor industry.

Chris Rowen of Tensilica went last. He said:

* The IP forecast is what is important.
* The scale of the IP industry is greater than $1 billion.
* IP is a success story.
* The growth is 20% to 25% per year.
* There are fundamental changes in the IP industry.
* EDA and then IP are fanning out.
* Now the industries are broader and cover manufacturing to software.
* There are two big ideas:
… IP and EDA merging
… Move to mastery of DSM and DFM to manage Moore's Law
* What is the role of design automation?
* It is moving to design creativity as block-level integration.
* But the picture is even bigger.
* The block is becoming the new NAND gate.
* The microprocessor is the new NAND gate.
* The aggregation of NAND gates is the new metric.
* It's better than Moore's Law.
* It's now the focal point of the EDA industry.
* The IP industry is healthy and growing.
* The fundamental trend is large processors as the focal point for design.

****************************

After everyone had their turn, there was a Q&A session with Q's from Jay, and A's from the panelists. Among the comments:

* Sanjay Srivastava said IP is important.
* Jim Douglas said R&D collaboration between customers and EDA vendors is important.
* Chris Rowen said designing at higher levels of integration is important.

Some discussion centered on industries adjacent to EDA that might provide opportunities for expanded markets. Among the comments:

* Wally Rhines said there are lots of opportunities in the automotive industry.
* Aart de Geus said yield and DFM are adjacencies whose time has come.
* Mike Fister said: "Can we demonstrate more value and ask our customers to pay for it. That will allow us to invest in the adjacencies."

************************

Then Jay said good night, and everybody applauded. Then everybody networked and then everybody went home.

****************************

Editor's Note: It might have seemed a little eerie to some to be spending time after hours in the fairly deserted HP headquarters. It's hard to miss the aura there, lingering from the recent demise of a high-profile HP CEO. But if anybody in attendance noticed – particularly the plethora of EDA CEOs on hand – they didn't talk about it.

Before the panel began, as people milled about with wine and hor'dourves, I stood and studied the portraits of William Hewlett and David Packard that hang discretely in the lobby of the building. I wondered what H&P might have thought about it all, but they didn't seem to want talk about it either.

Perhaps they were thinking the press that covers HP only repeats rumor and gossip, and what they should be focusing on is the customers. Maybe they were just thinking about oscilloscopes.

Who knows?

****************************

February 28, 2005

Peggy Aycinena owns and operates EDA Confidential. She can be reached at peggy@aycinena.com

Copyright (c) 2005, Peggy Aycinena. All rights reserved.