EDA is DEAD
Years ago I bought my ancestral home, the house where my beloved grandparents lived, the place I grew up in. It was more an emotional investment than a financial one, much more. After completely renovating it with my keyboard hardened hands, reliving much of my childhood, I joined the ranks of the slum lords and rented the house out to complete strangers. Thanks to ungrateful renters, inconsiderate neighbors, and a vindictive housing inspector, there are no thrills left whatsoever, just tedium and frustration. Blogging about EDA is much the same, the roller coaster excitement is pretty much over, now it is more of a carousel ride, so yes the EDA I once knew is dead.
EDA really came into it’s own with the advent of the fabless semiconductor manufacturing model pioneered by companies like VLSI Technology and LSI Logic. Using excess Japanese manufacturing capacity and building its own fabs locally, they built Application Specific Integrated Circuits (ASICs) and offered ASIC services for emerging fables semiconductor companies. A few years later TSMC introduced the pure-play foundry model promoting manufacturing efficiencies and commercially available EDA software. At that time semiconductor manufacturing processes were very different and could be exploited for competitive gains, re-usable Semiconductor IP was in its infancy, so the value proposition of EDA software was clear and present. Today, unfortunately, that is no longer the case. Commercial Semiconductor IP dominates the area of an ASIC and designs can be moved to second and third source foundry partners with little or no change. The biggest design challenge now is adapting to the new process geometries which requires hands-on experience.
The fabless ASIC Services business re-emerged in the year 2000 starting with eSilicon, followed by VeriSilicon, Alchip, Open-Silicon and a whole host of others. It has been a long, hard road with fierce competition (even coming from the foundries themselves with Global Unichip and Faraday). Honestly I did not see the long term value proposition back in 2000 but clearly it is here today. ASIC services can cut total costs by one third and significantly reduce the internal risk of designing to a new geometry. You can also minimize front end expense (IP/NRE) and back load the cost on a per packaged chip pricing agreement.
So where does that leave EDA? Venture Capital has left the semiconductor, IP and EDA market segments so budgets are a fraction of what they once were. ASIC design starts are rapidly declining, EDA innovation is stalled, and ASIC Services companies are multiplying like rabbits. What else are unemployed ASIC designers going to do, work at Starbucks? Not to mention a smattering of foundry created EDA tools now hitting designer’s desks, which is understandable. As manufacturing processes become commodity, foundries will use proprietary EDA software and design enablement services to support their wafer value proposition. Bottom line: the ASIC Services market is $1B+ today and projected to exceed $10B by 2015, EDA on the other hand is a flat $4B industry that is still living in the house it grew up in.
How does the EDA revenue curve look compared to TSMC? I don’t think you can declare something dead until you can show what has obsoleted or replaced it. By the way, Venture Capital is down about 50% across the board according to the PWC MoneyTree survey for Q1/09. I think you need to look at more data and gather some more context before you say EDA is dead.
If blogging about EDA is tedium blog about what you are passionate about. You will certainly find it more energizing and you may very well find a new audience that shares your passion.
Sean,
Please re-read, I qualified the EDA is DEAD title as EDA as I once knew it is DEAD. EDA will certainly never go away but it is a flat $4B industry and has little chance to grow beyond that unless significant changes are made.
If you read my entire blog you will see that I am actually very pro EDA, I want EDA growth to track semiconductor growth. Seriously, is that too much to ask for?
I’m an EDA/IP business consultant, my customers want their businesses to grow, that is what I do, that is the whole point of this blog, I want EDA/IP to grow with the semiconductor industry.
D.A.N.
Dan
As an EDA headhunter that started in 1984 when Daisy, Mentor, and Valid were just getting started and have spent half my life in this industry………………………YOU ARE CORRECT SIR.
The mystery to me always was how an industry that invented something that I consider magical never really got paid what they should have while their customers grew to bb$$ valuations and couldn’t have done it without the tools.
But what do I know
Rich
Rich and Dan,
A year or two ago, there was a mock documentary called ‘A Day without a Mexican’ to outline the role of latinos (some possibly illegal) in low wage labor. Maybe the EDA industry can launch a PR campaign called A Day Without EDA to emphasize the contributions of CAD to the electronics design industry!
Imagine one scene where a couple of chip designers attempt to transcribe a gate layout on graph paper into an Excel spreadsheet, coloring sets of columns and rows … (can I claim the copyright on that scene?)
EDA is on life support for business reasons. EDA executives have been semi-professional. The last EDA CEO with a head for business strategy was Jim Hammock. Since he moved on, we have seen only tactics, reaction and fantasies of better mousetraps.
As Rich points out, EDA never grasped its own value strongly enough to cash in on it, EDA customers, on the other hand, are serious business people who have thoroughly punked the semi-pros.
Ask any EDA startup CEO what his business vision is, and your question will not be understood. You will get something about technology. A tough customer diverts that conversation to cost reduction, which leads directly into the current downward spiral.
Dan, I agree with you 100% that EDA is dead. I used to ardently like EDA, and always amazed by the way algorithms interplay with chip design. I was interested in electronics and chip design, right from my school days. When I was in India, I discovered EDA and wanted to work in this field right from that time. After finishing my graduate school and working in chip industry for 6 years, I got my chance. I was so eager to work in the industry that I did not care the salary or benefits. But once I was in, that is when I found the harsh truth. The industry was on its way down, with protectionism was on the core. Friends didn’t talk to each other because they may ask questions about the job. The sales was all done based on relationship with no competitive advantage for any company. Also, customers continuously want lower prices and closing accounts is hard as hell. Everyone is bent on undercutting each other. Now after spending several years in this EDA industry, where these habits have completely set, I am out of the industry. But I still genuinely feel for the EDA industry. Whatever changes need to be made, I feel that the biggest change should be in the value delivered. EDA also needs to take more ownership of the design. With # of designs going down rapidly, EDA should realize that the value has to be provided to fewer and fewer customers and it needs to deliver specialized solutions for these companies. You can be at $4B valuation for it may be the lowest point. But if the industry wants to grow, it needs to make more bolder decisions. I hope the leaders realize that the time is now and the weaker companies should give way for more stronger companies.
Dick Lanham,
I’ll argue that Jim Hammock had a flawed view of business strategy because I worked at Silicon Compilers when Jim sold it to Mentor Graphics (for a 5 to 1 reverse split, ouch). We had totally overlapping products for:
1) Schematic capture (Design Architect won)
2) IC layout (IC Station won)
3) Place & Route (IC Station won)
4) HDL Simulation (QuickHDL won)
5) VHDL Simulation (Modelsim won)
As you can see, the Mentor classic tools all won while all of the SCS tools (GDT, Genesil, Caeco) were eventually killed in spite of their superior features and technology.
Out of the 400+ people from SCS at the time of the merger, there are only a handful left at Mentor today.
That is not a successful business strategy, rather a failed strategy.
Daniel
Dan,
Using such an inflamatory title as “EDA is DEAD” succeeded in drawing me into conversation — let me respond to that simple thought first. Here’s the problem: there are only a few companies that write/create the hideously complex software to address our engineering applications, and those companies have a rather small customer base: no Joe user is picking up a copy of Synopsys VCS from Amazon.com. Despite the pain in our engineering budgets, we cannot do our jobs without the proper tools. So, no, EDA is not dead as long as more than a couple of companies are creating chips.
There is also a bias in your evaluation toward ASICs/ASSPs. The new upsurge in FPGA design complexity is putting some new life into the “do-it-yourself” designing — the kind of excitement I remember from the long-ago ASIC design days. Although the IP/Specialist design firms are gearing up to address the FPGA market as well, there is still a lot of the old wild-west designing going on in this segment that will probably continue for some time, owed mostly to the reprogrammability of the parts and their new, respectable performance and capacity.
As far as excitement, though, in the EDA industry, I will have to agree with you. It just doesn’t seem like people are terribly interested in creating or working on new technology in this field, and it shows. The tools that I use still look a lot like the ones I used over a decade ago — the promise of powerful and easy to use tools has never been fulfilled. The tools and their use is still as arcane as ever, perhaps even more so.
Let’s hope that necessity and innovation will meet and that EDA can regain some of the luster (and fun) it once had.
What’s exciting about EDA nowadays is the ability to create new tools on a showstring. With the abundance of free libraries ,algorithms and programming languages, a small team can create tools that would have taken many hundreds of people working for years not long ago.
Now THAT’s exciting
Cheers
Disclaimer: Altium is in the business of providing electronics design tools. Note, I didn’t say we are an EDA company – for a number of reasons. First, it strikes me that “EDA” remains a confused term, sometimes meaning board-level design, sometimes meaning IC and silicon design. We prefer “electronics design” and we’re quite happy not to be pigeonholed in a market definition of EDA – because of what Dan says. If electronics designers are to thrive, they will need to change how they design, what they design, and for whom they design, all in the near future. So their tools will need to change (or, at least, their choice of tools). And so EDA will also need to change.
We think electronics designers will change and can change, and of course we’re in the business of helping them. With perhaps 15 billion internet connections predicted by about the middle of the next decade, someone will have to design the gizmos that connect, and they won’t just be computers and comms devices (of whatever sort). The users, including of industrial electronic products, will simply demand this connectivity. Electronics design will move into the realm of “connected everything”. The tools used by designers will therefore have to provide more focus on facilitating the design of the intelligence in the product, and how it connects – both without compelling the designer to reskill. Those of us on the tool side that work that out will, we think, have a future. We’re seeing evidence from users that they will change (themselves and their tools) – a good sign for the industry, and for the EDA sector – of those bits of it up to the challenge.
Hi Dan, your post is incomplete without speculation as to what these significant changes could be.
Marc
Marc,
I have referenced business model changes for EDA in various blogs. Most recently in http://danielnenni.com/2009/11/02/tsmc-oip-explained/
In comparing EDA, Foundry, IP, FPGA, and Design Services.
The success based model where the vendor gets the bulk of payment upon successful completion of the design via (tape-out fee) and/or silicon royalties. Versus EDA where you pay money upfront and have little invested in the success of the project.
How big would the FPGA market be if it cost $1M for tools?
The foundry margins on FPGA silicon are very small in comparison to ASICs, so the decline in EDA/ASIC design starts hurts us all.
D.A.N.
I’m not a veteran as other people who commented here. But I agree with you Dan, the traditional EDA is dead! The problem is that the EDA companies are using the traditional business model of software companies. Developing software that satisfies “general needs” and customization for individual customers. This customization is either done by the EDA company itself or by customer themselves. However, the EDA market is special because the user of the software are pros in the field. They are usually the EDA department in the design houses, and they worked on the particular field for years. So they know exactly what they want and they know exactly how they can get what they want. As the result, the development team of EDA company is constantly asked for special customization and eventually comes up with a huge pile of features which has to use a very complex sales model to get the money. Effectively, EDA companies are fighting the trend of customization in some way instead of embracing it. IMHO, customization should be in the heart of EDA company’s R&D effort. It should change its focus from big unified flow or platform into flexible customizable feature offering, and so its products should be marketed and sold this way.
A great essay, Min! Your analogy on the EDA business model might be further qualified to that of enterprise software companies, such as SAP or CA. However, EDA is more specialized to maybe a couple dozen large enterprises worldwide but many, many SME’s (small to medium size enterprises). It is not general software that is scalable, so it is difficult to do the same with pricing and that creates difficulties for customers who are SME’s including fabless chip start-ups.
Providing customers with the capability to do their own customization, rather than put that burden on the EDA supplier (and secret keys to unlock custom features), brings up an issue on the customer side: are they willing to provide their own resources to carry out such customization (scripts, plug-ins, IP-XACT generators using that standard’s TGI, etc.)?
Dan, Exceptions to every rule: Atoptech is finding a seam in a flat to declining market. Smart,experienced teams committed to innovation find a way.
Dan is right about all this. EDA is Dead.
I’ve worked for a couple of different semiconductor companies over the years and have seen several trends that work against the EDA business.
#1 FPGAs have killed ASIC starts on the low end
#2 many ASICs have been replaced by general purpose CPUs with some well-tuned firmware
#3 the old tools from the 1990s were good enough. why should i buy new ones (or different one).
I think there is always opportunity for Design Automation tools, but in the chip business, EDA is pretty much a solved problem. So its time for these EDA companies to move on and look at other industries that could benefit from design automation tools.
Not dead, just moribund. I would say that EDA has not been adapting to changes in technology and economics (that have been fairly predictable), and has failed to innovate. At this point a bottom-up rebuild of the tools is possibly required, e.g. most of the tools from the 90’s know little about power management, multicore or DFM.
I suspect major consolidation is in the offing, and if EDA tool availability/pricing is getting in the way of building working Silicon I think the actual Silicon-vendors may step in to provide the tools (as the FPGA vendors do).
Hi, Kevin! The great recession of the past two years, and the so-called recovery from the 2001 tech crash (at least for high tech) that crawled from 2003 through 2007, has not helped EDA nor (more importantly) its customers, the systems and chip houses. Start-ups are the main source of innovation in EDA but VC has been frozen in place, affecting EDA start-ups worse than fabless semis. Two successful acquisitions that happened in this decade (I’m sure there’s a couple of others) were Synopsys taking Synplicity on a late stage exit and Mentor Graphics taking Lighthouse (formerly Silicon Forest), creators of the inFact verification technology. I fear that the dry pipeline for EDA may be even drier as we wake up in a recovery.
Hi Dan,
You are right EDA is dead. Today no VC in the valley will fund a new EDA company and it is highly unlikely that they will fund a semiconductor company.
Teams get by using old methodologies and old tools. Synopsys adds hacks to VCS to generate incremental revenue from customers.
Younger engineers won’t touch EDA and where are the new ASIC designers. Certainly no in the Vally.
Where are the university professors that are championing new tools? Not at Stanford or Berkley.
This industry was killed by sales and marketing machines which were out negoiated by semi company business people who had the likes of Andy Grove lomming over their shoulders demanding that they drive a hard bargain.
Who are the real losers?
The engineers and the semi companies. You get what you pay for. They will pay for not supprting the EDA industry with lower productivity using outdated tools and methodologies. What do the companies do instead of supprting innovation in the EDA space? They ratchet up the pressure on the engineers and demand longer hours and destroy lives.
I agree 100%. Silicon waste will continue and it’s the consumers that pay the ultimate price.
D.A.N.