Dialogue with Rajeev Madhavan
by Nitin Dahad
With increasing consolidation facing the EDA (electronic design automation) sector, it is worth remembering how much of the EDA innovation used to be driven by new start-ups. Here, you will be able to read again the interview that we carried out with Rajeev Madhavan in 2006 (courtesy of The Chilli RED, our premium content service), hoping it will inspire more entrepreneurs to contemplate starting out on their own and break out of the EDA logjam.
Rajeev Madhavan is chief executive officer, president and chairman of the board of directors of Magma Design Automation, the company he co-founded in 1997. Magma was Rajeev’s third start-up, after previously starting up and selling off Ambit Design Systems, and before that LogicVision. The Chilli met up with him in Silicon Valley to listen to his experiences and insights into being an entrepreneur in the EDA industry and some of his perceived challenges.
Rajeev Madhavan
The sections/questions in this interview
- Rajeev Madhavan's background
- The start-up vision: can you tell us a bit about what drove you to do a start-up, and some of the issues leading up to it, including raising money and who influenced you?
- How did you get your first deal? And how did you respond?
- How did you build your team, how did this evolve, and how would this compare to the industry today?
- What drove the team and what would be your advice to other start-up management teams?
- What’s your advice on dealing with VCs and raising money?
- What are the main industry challenges today from your viewpoint, and what are the opportunities?
- How do you stay on top of tracking developments and the company?
- What are the company challenges you face now and could you put this in the perspective of the industry challenges?
- Can you highlight some of the sacrifices that you made in starting up and growing successful businesses?
- What have been your toughest and proudest business moments?
- What do you enjoy and what do you hate?
- What would you do differently?
- Did you ever feel like giving up?
Background
After receiving his bachelor's degree in electronics and communication from KREC (Karnataka Regional Engineering College), Surathkal, India, Rajeev Madhavan went on to study for a master's degree in electrical engineering at Queen's University, Ontario, Canada.
His first job after graduation was at BNR (Bell North Research) in Ottawa, Canada, where he got an opportunity to write his own synthesis tool to complete a chip design he was working on. He joined Cadence as a full-time employee in 1991, and worked for just over 12 months on Cadence's Spectre HDL tool.
Troubled by the politics of getting new tools into customers’ hands, he left in 1992 to set up LogicVision along with Vinod Agarwal and Michael Howells. Here he felt he could do better than the synthesis tools he was using so in 1994 decided to set up on his own and founded Ambit Design Systems. Having achieved some significant benchmark performance against prevailing Synopsys tools, Ambit was eventually successfully sold to Cadence.
Rajeev Madhavan continued to be driven by the need to deliver what the customer wants, so founded Magma Design Automation in 1997. Magma successfully raised several rounds of VC funding (over $100m), and went on to become a public company in November 2001.
The start-up vision: can you tell us a bit about what drove you to do a start-up, and some of the issues leading up to it, including raising money and who influenced you?
Magma is my third start-up, after LogicVision and Ambit Design. It was Karen Vahtra, whom I met at Ambit, that convinced me to do it a second time – she was quite pushy, and would do whatever she felt on the spur of the moment. She was the ambitious one, I was the cautious one.
For the third time round, I called Andy Bechtolsheim, the co-founder of Sun Microsystems, and told him, ‘This is the problem I’m going to solve (about which I was very clear)’, and asked for an investment of $1m. He offered $2m but we only took $1.5m from him.
I had the idea in February, and ‘saw’ a few prospective customers to take a look at their chip problem. Andy had a chip going through similar problems at Cisco so he understood the issue.
Karen Vahtra was one of the Magma co-founders. The company was funded within 45 minutes, and incorporated on April 1, 1997. The idea was to get a prototype done before raising any money, with our goal being to complete it within one year. There was another competitor at the time, Monterey Design Systems. They were nine months ahead, so my natural instinct was that I had to beat them. Every time I talked to my people I’d tell them that they were nine months behind schedule so that would get them working harder to achieve our milestone.
Key people during the start-up experience with Magma were Mark Perry, a general partner at New Enterprise Associates and ex-CFO of SGI (Silicon Graphics), and Andy Bechtolsheim. Mark is excellent to work with and he kept me on the right path all along, making sure that important things such as keeping accounting records were done well. Andy falls under the genius category in terms of depth of knowledge and understanding of both the technology and the business.
How did you get your first deal? And how did you respond?
In the early days, people were holding us back with legal threats – but we weren’t going to be held back. In 1998 we set a target to produce a demo, and in 1999 we showed it at DAC (the Design Automation Conference).
By the end of 1999 we’d booked our first deal. The CEO of 3DLabs, Osman Kent (now a musician), called us because he’d asked Monterey who their competition was and that’s how he became aware of us.
So we bet the company on a 10 million gate chip for 3DLabs, which by tape out became 13 million gates. That was the stress test we needed: I thought, ‘If I mess up, 3DLabs could die: I couldn’t bear that thought’. At the time when we were working on this chip, everything else was put on hold, with the whole company working on the 3DLabs chip. Some people actually thought we were a part of 3DLabs.
While working on this, the entire R&D team was responsible for making the code happen.
Our sales team had also closed other deals during this period, but we had to put a hold on these as we were focused on delivering for 3DLabs.
How did you build your team, how did this evolve, and how would this compare to the industry today?
We held extremely tough interviews for the first 40 R&D staff – the interviews were held over 1 to 2 days and we looked for code-writing capability and technical talent. We were getting very stellar people and we had created a feeling in the industry that Magma was the company to work for.
During the first 18 months everyone reported to me – around 40-48 people. On Wednesdays and Fridays we had a team meeting and everyone had to give a two-slide presentation.
In the early days at Magma, money was not an issue and any funding round was not a problem. Every round was over subscribed.
That is different to today’s market, in which you need to get to a prototype stage much more quickly but with a lot less money.
What drove the team and what would be your advice to other start-up management teams?
We were always driven by the philosophy that the customer is always right – so we’d chase what was needed and get it done. However, we were not shy, and if money was needed to get the job done we’d ask for it.
If you don’t have the business/commercial talent, then as an entrepreneur you had better get someone with those skills quickly. In addition, a key attribute for entrepreneurs should also be a willingness to take advice – this is very important. On this latter example, Magma has invested in two un-named start-ups, one of which is willing to listen to advice that will help their business, while the other thinks that they know it all and therefore don’t need advice. Needless to say, the start-up which has followed advice is doing better.
It is important to know when to call in advice and when not.
What’s your advice on dealing with VCs and raising money?
My advice on dealing with VCs is to be very up front with them, and you need to be clear in your mind what they should be capable of doing for your business. It’s also important in this aspect to know and define what you are good at and what you’re not good at.
Face-to-face time is vital, but no amount of prodding by the VC should allow you to give in and allow them into the operational side of the business.
Other useful tips related to money raising:
- it’s prudent to raise money earlier than when you need it
- don’t wait until you have reached your last drop
- don’t put the company at risk.
What are the main industry challenges today from your viewpoint, and what are the opportunities?
The consumerization of semiconductors means a number of challenges, consisting mainly of:
- power (low power consumption requirement) changes
- the end user focus and consumer applications driving the industry
- the drive toward smart communications and smart appliances
This is a significant change to the past, in which the computer industry was the main driver for the EDA business.
In addition the emergence of India and China will mean double or quadruple the number of new opportunities. Although there is very little infrastructure in India, there is a lot of opportunity, as many graduates engineers will be hired by the semiconductor industry from there.
There are two ways of looking at the challenges posed by the globalization of the work force. One is to say it will never happen and look away; or the other is to use these challenges as opportunities. We have to treat the economy as a global one, so we take advantage of the global workforce.
How do you stay on top of tracking developments and the company?
These are the main ways in which I do this:
- use my blackberry/PDA
- walk round the building and spend time in other people’s cubes (cubicles)
The latter gives me a direct understanding from the team – which sometimes helps me to relate to their issues.
The reason I take this approach is because you need to have your feet on the ground. You also need to have people who don’t mind their CEO walking down and talking to them. By building a culture like this within the organization where you can raise the issues, it easier for the management team to take action accordingly [to improve the business operations].
The hard part is to clearly define my job and not get involved in everything: I would be a very lousy VC investor because I would be in the invested company’s face all the time.
What are the company challenges you face now and could you put this in the perspective of the industry challenges?
The company is now undergoing major change – from being ad-hoc to being more structured and organized.
We have become the target for many start-ups, in terms of competition: many want to be the “Magma of EDA”: some we fight directly, others we take on as partners via appropriate mechanisms and funding. I encourage other start-ups in the EDA space. I disagree with leading EDA figures who suggest there is room for only three major companies in EDA. My take is: yes there are three major ones at any given time, but the group making up those three has evolved and changed over time.
For example, in the first 10 years we had Calma, Applicon and Daisy; in the second 10 years Daisy, Mentor and Valid took over the mantle for the top 3; and in the third decade it was Synopsys, Mentor and Cadence.
I believe there will be displacements in this fourth decade, as we are going through those changes now.
In terms of attitudes of the EDA industry, I agree with the Philips CTO who during his keynote at DATE stated that it is the responsibility of the EDA vendor to create the solution for the customer – not to provide just the niche tools, but also provide the integration and automation of the flow.
The problem today is that EDA vendors provide only niche tools or point tools – and that is wrong. The overall cost of doing chips today has gone up, but EDA as part of the overall semiconductor spending has gone from 2.1 percent to 1.7 percent. The customer’s productivity depends on the level of automation the EDA vendors provide. This is the paramount problem that the industry needs to address.
Can you highlight some of the sacrifices that you made in starting up and growing successful businesses?
My family has been tremendously supportive, to the extent that my daughters think of my competitors as their competitors. For example, when seven years old, one of them said at a party, “you shouldn’t be talking to her Mummy”, [because she worked for a competitor]. Both my daughters are very interested in business and are indeed very competitive.
In terms of sacrifices, my main sacrifices were at Ambit, where I had lots of debt and a huge amount of financial risk. My wife worked and brought home the money during this period – what she has done to support me and achieved has been phenomenal.
I have indeed sacrificed time with the family [in doing the start-ups], but now I am spending a bit more time (compared to the early Ambit/Magma days) with my family.
What have been your toughest and proudest business moments?
My toughest business moments were the fights we had on the board at Ambit just before I formed Magma.
My proudest moments were:
- the final tapeout of the 3DLabs chip
- IPO in 2001 (against the tough economic climate for IPOs)
- When Ambit Design Systems was bought by Cadence. Prior to the acquisition I was worried that Ambit would fail, so to see a good exit made me happy, and the timing was perfect.
What do you enjoy and what do you hate?
What do I enjoy? Well EDA is a very geeky world in which you need hardware and software experience, and the customer base is also highly talented. So I enjoy walking around and talking to people in the industry, both internally and externally.
What do I hate? The travel. I despise doing so many trips – doing six weeks a quarter on a plane gets you to the point where the flight attendants know me by name.
In addition, the EDA industry is very myopic, and vendors are often focused on their own problems, resulting in customers having to spend on unnecessary things. Why can’t vendors get it right? We are destroying the value of the industry.
What would you do differently
?Every start-up is a learning experience, based on making 80 percent good decisions and 20 percent bad decisions. I try and learn from the 20 percent and make them better the next time – I actually make a mental list of those things and move forward.
At Magma, we are actually very good at this kind of self-analysis.
Did you ever feel like giving up?
The main point is that we align ourselves with what the customer wants and that helps us to win.
The reason I don’t give up is because I’m very competitive and believe that the EDA industry isn’t providing the automation that designers need. Increasing automation is Magma’s goal.
Magma’s competitors also keep me very focused and motivated because they often use unfair methods. Every time I begin to think about moving on from this industry, their unfair tactics drive me back into the ring to teach them a lesson.
Comments on this story? Send an email to the editor: address - editor [@] thechilli.com
 |
|
|
© Chilli Publishing Ltd 2008 |
22 JUN 2008 |
