Q: Whither EDA? A: Private Equity?

Welcome to DAC week! I'm not there, but I expect to see a flurry of announcements and will be looking to keep up with things in EE Times.

This morning's paper has a story about "private equity" firms being interested in investing in Cadence: Cadence would offer buyer steady cash flow. And, an investing blog has Cadence Design Negotiates With Buyout Firms.

This is all at the rumor stage, but it's interesting in what it says about EDA. Rather than being a high-growth industry where one might make several times investment by picking the right hot technology, this is more about lucrative cash flows coming from software maintenance agreements.

If private equity firms become players in our industry, how will this change them? Will it lead to "slash and burn" of R&D, just trying to milk the recurring revenue from existing products? But how could such an approach fund the R&D to keep pace with the changes and challenges following from Moore's Law?

Update: An EE Times blog posting on dynamics of the EDA business. And, a YouTube interview with Gary Smith and his DAC preview. EDA gets hip!

More Updates: Richard Goering weighs in.

DAC Linkfest

Last year, I went to DAC in San Francisco, and I made a number of posts to my blog both before and after the 2006 conference:

• DAC is Here!
• Gary Smith's Must-See List
• Cooley's Must See List for DAC 2006
• 2006 DAC Report

This year, I'm not going to DAC in San Diego, but I thought I'd at least "phone it in" for my blog. With due credit to the EE Times' Richard Goering and other professional journalists, here's your one-stop 2007 DAC blog nexus:

• What's coming up for DAC. Less DFM & ESL hype, more functional verification, STA, and low-power.
• 40 new exhibitors at DAC
• Missing in action at DAC 2007
• A "what's hot" list for DAC 2007
• DAC & DFM – Once More, with Feeling I can't tell who the author of this article is. But judging by the site and the length of the article, I think it must be "EDA Penny".
• John Cooley's DAC Cheesy Must-See List. John hasn't posted his list as of this writing, but it should appear soon at his DeepChip web site.

Existentialist EDA

I'm getting tired (so tired! --Bauhaus) of waiting on vendors who promise that all our problems with their product will be fixed in the next release!

I told an EDA vendor that waiting for their tool to improve reminded me of the play Waiting for Godot. I thought this quite clever, but they didn't seem to. Maybe they didn't have to read it in college?

It might be fun to cast EDA persons into the play. Who would play the roles?

Vladimir

The tramp who struts about stiff-legged due to a painful bladder condition.

Estragon

The tramp who complains of his ill-fitting boots

Pozzo

cruel but lyrically gifted man

Lucky

servant

The Boy

messenger

Godot

?

Here's my literary interpretation:

• Clearly, Vladimir and Estragon are the afflicted customers.
• Pozzo is either the sales or marketing guy (I often forget the difference).
• Lucky is the poor FAE.
• The Boy is post-sales support.
• Godot is ??? Hey, I don't have all the answers, use your imagination!

Something Completely Different (but Worthy)

Dear readers,

Off topic, though perhaps a reason for the decline in my posting frequency since the beginning of the year. For the past several months, I've been supporting a program that raises awareness and funds for The Leukemia and Lymphoma Society. This charity funds blood cancer research and patient services.

The program I'm part of is called Team in Training. Every season of every year, hundreds of runners, cyclists, and triathletes come together to train for a signature endurance event (e.g., marathon, century ride, or triathlon). More importantly, we collectively raise hundreds of thousands of dollars to help find a cure for cancer and to help those currently afflicted. We do this by getting sponsorship donations to the Society from our friends, families, and colleagues. Why not give blog readers a chance to help, too? That's where you come in. If your curiosity is piqued, and you'd like to read the rest of the story and perhaps make a donation to a very worth charity, please take a look at my Team in Training page. There you will find a link to another blog I've been keeping, "New Rider on the Purple Stage", which describes my cycling season in pictures and words. Please note that my season culminates in a few weeks, and the last day for donations for this season is May 21, 2007.

Thanks for reading.

John

Executive Blogging

As seen on Dilbert.com:

Pretty funny, but hopefully not prevalent in our industry. For example, I recently started reading Jonathan's Blog by the CEO of Sun Microsystems, and it's pretty substantive and useful.

What's Special About This Number?

What's Special About This Number? was featured in IEEE Spectrum. It's an interesting list of factoids for past, present and future fans of mathematics.

Enjoy!

"How-to" SPICE Book

It's been several years since I've run SPICE myself, but I'm finding it increasingly needed even for a RTL/gate-level engineer. Could anyone recommend some really good SPICE books, which aren't just a reference but show you how to do things? You can post your recommendations and advice here as comments. The "things" I'd be most interested in doing would be characterizing digital standard cells and analyzing IC interconnect.

I looked at Amazon.com and didn't find a lot of books like this (no "SPICE for Dummies"?), but these two look interesting. Before I or my company spends a chunk of change, do you know these titles and how useful they are?

Legend of Test

Peggy Aycinena did a fine interview with Tom Williams, truly one of the giants in the field of Test. The guy invented full-scan DFT in the 1970s! Can you believe that? It's remarkable that his career spans 40 years, and he's been productive throughout. He's published many seminal Test papers, and has received the "Best Paper" award at DAC.

Be sure to go to the last page and see a list of his self-selected most important papers.

Hither ASIC

q.v. my earlier posting Whither ASIC, EE Times just ran a story on the declining activity in the ASIC business.

It's a thorough survey of what's happening in the business. I agree that designs need to see more IP reuse -- the extreme case is to use an ASSP. And I'm skeptical of Structured ASIC, and it seems the market has also given that idea the thumbs down. But the article didn't mention the trend of more fabless companies finding ways to do COT rather than ASIC design, possibly through "virtual ASIC vendors" like eSilicon or Open Silicon.

Intel to Build Fab in China

"Our goal in China is to support a transition from 'manufactured in China' to 'innovated in China'"
--Intel CEO Paul Otellini

Ah, globalization!

Whither ASIC

I'm surprised by how bad of shape the traditional ASIC business is in. Are any companies making money? Only IBM Microelectronics?

Several of the Japanese semiconductor powerhouses are cutting way back (or closing) their US ASIC engineering. Today a friend told me that LSI Logic isn't doing ASICs any more! Is that right?

The vendors that are doing the best have attractive IP portfolios, like IBM or TI. Those that historically relied on manufacturing process (e.g., Japanese) and left IP entirely up to the customer aren't faring well.

I'm especially surprised that even fabless chip startups aren't using ASIC vendors! They invest in the COT model or the "virtual ASIC" model, contracting with a number of service providers. I would have thought that the integration value-add of ASIC vendors would be even more attractive in these days of worrying about signal integrity, variability and DFM. Don't chip designers want "one throat to choke" when things going wrong?

We may not have ASIC vendors to kick around for long.

Britney Spears guide to Semiconductor Physics

Here's some fun on a Friday. EE Times gave me the link to this crucial technical reference, Britney Spears guide to Semiconductor Physics - The physics of optoelectronic technology . The material is technically sound and more accessible than a textbook.

Check it out if you don't mind gratuitous pictures of Ms. Spears. What's next, Paris Hilton's analysis of the Traveling Salesman Problem? (a Hilton ought to help with that ;-)

Immortal, Indeed

Those of you who read the print edition of EE Times may recall a funky little contest/feature that ran at the back of the paper, "Immortal Works". I think I can remember this *always* being part of the paper, even going back to the 1980s. Here's the latest version I could find.

Last month, EE Times went through a redesign. There are many changes to the organization and layout of the paper. It's actually printed on smaller (shorter and narrower) paper than before, in line with several major mainstream papers. More cost reductions from the ink and paper crowd, probably. I actually like the smaller format -- it's easier to carry around and peruse at opportune moments.

But I won't like if Immortal Works has been canned! It's missing from the redesigned print edition! According to the editors' explanation The new EE Times: more than a redesign:

Taking a hiatus will be Immortal Works, our longtime caption-writing contest of which many of you are devoted fans. We're moving the feature online, where the audience interaction is fast, efficient and easy.

Like I said, it was a sort of goofy feature. And even though I never had the time (or more likely, the wit) to submit an entry, I miss this quirky part of EE history! Immortal Works, where art thou?

Standards? EDA has lots of 'em!

While I'm pleased that the UPF power standard plans to converge with the rival Common Power Format (CPF), it begs the larger question: why should we have multiple standards organizations, anyway? I'm referring to Accelera, Si2, and the IEEE. It seems just the IEEE should be enough.

It's like professional boxing, where there are now so many sanctioning organizations (WBA, WBC, WBO, IBF, ...) that the average person has no idea who the real champions are, and no longer really cares, either.

CAD Engineer's Bookshelf

Inspired by the DFT Bookcase over at the DFT Digest blog, I decided to create pointers to the technical books I use every day in my work. (Yes, I looked the ones strewn about my desk and next to the computer, rather than those arranged neatly (and less used) on my bookshelf. Perhaps you'll discover a title or two that will help you.

	Great book for learning Perl -- it's how I learned, going through it from cover to cover. And I still use it for a basic reference.
	A surprisingly useful book. I can't count the number of times that I needed to solve some puzzle in Perl, and found it already grokked in this book.
	Lots of good ideas here. When you are going to pick a coding style, why not follow what they've already figured out?
	My newest addition. I'm enjoying working through it to learn about references, packages, and object-oriented Perl.
	Programming Perl is the "bible", though personally it's not my favorite. And I find Larry Wall's humour distracting.
	My go-to book for Tcl.
	Because, sometimes, a good ol' shell script is enough.

Ambric Multi-Core Design Wins Award

I like this press release: Ambric Wins In-Stat's 2006 Microprocessor Report Analysts' Choice Award for Innovation because of two eye-catching features:

• the claim that this multi-core design was first architected for easy programming, before creating the hardware architecture. I agree with the claim that "most massively-parallel chip companies started with the hardware, leaving the programming model almost as an afterthought".
• the implementation is globally asynchronous, locally synchronous (GALS). Man, that is sexy! ;-) I've been interested of commercial implementations of this more-efficient clocking approach.

I'm going to watch for Ambric and try to learn more about their designs. By the way, interesting caveat in the quotes: "If Ambric's tools work as well as the company promises, ..." Ain't that always the challenge?

That’s OK, the Corporate Raider took my backpack

Jonathan David is an active IEEE volunteer and he's the local chair of PACE, the career-oriented part of IEEE. He's written a nice post on career planning, Professions of a Silicon Valley Double-E: That’s OK, the Corporate Raider took my backpack. [on keeping your parachute ready].

It's a nice short read. One of the "takeaways" from his advice is that he highly recommends this book, which is surely a classic in career planning:

A Taste of DesignCon

I didn't go to DesignCon, but Gabe did, so let's sample his DesignCon Morsels.

What's nice is the variety of topics he touches on:

• The lack of interest in ESL vs. design implementation, and why engineers are being short-sighted by not getting into architecture.
• Thermal performance, an interdisciplinary problem. When I've been involved in thermal analysis, the biggest problem is coming up with a credible power consumption figure. It's back to the "vectors" problem, and what's a representative (or worst-case) simulation?
• Is DFM for Designers?
• Fewer students are going into Engineering. And engineers "get no respect", as Rodney Dangerfield would say.

Thanks Gabe, for the synopsis!

Dr. Thomas W. Williams to Receive EDAA Lifetime Achievement Award

Saw this over at DFT Digest: Synopsys Fellow Dr. Thomas W. Williams to Receive EDAA Lifetime Achievement Award

Look at all Dr. Williams' achievements, stretching back to the 1970s. This is (one of) the father of LSSD, which begat the full scan DFT methodology that we all use today. He's truly a giant in DFT one of the legendary early figures in EDA.

Congratulations on a well-deserved honor!

Chips Grab Mainstream Headlines

I was surprised and excited to see a chip-process development featured on the front page, above the fold, in the San Jose Mercury News. This is "the newspaper of silicon valley", as they say. The headline gushed Chip advance biggest in 40 years.

It's about successes announced by Intel and IBM to fabricate metal-gate transistors with "high-K" dielectric. This is good news, which hopefully will enable us to continue to follow Moore's Law.

But it doesn't hit me that this is the biggest development in 40 years. Is this triumph over materials significantly different from figuring out how to to use Copper interconnect, or "low-K" interconnect isolation? Those were both hard manufacturing problems that the process wizards solved in the last decade. Maybe it was a slow news day. ;-)