Peggy Aycinena is a freelance journalist and Editor of EDA Confidential at www.aycinena.com. She can be reached at peggy at aycinena dot com.
Warren Savage: the revolutionary in our midst
August 23rd, 2012 by Peggy Aycinena
Behind Warren Savage’s calm and courteous demeanor beats the heart of a revolutionary: A guy who not only talks the talk, but walks the walk of growing his beloved IP industry through the most radical of ideas – cooperation.
Warren is the founder and CEO of IPextreme, a Silicon-Valley based company helping other companies commercialize their IP, small nuggets of pure gold that would otherwise enjoy only internal use. With the assist of Warren & Co, that gold is beefed up, intensely documented, and then licensed to users outside the firewall who then have access to robust 3rd-party design blocks, yielding revenue back to the IP developers they would not otherwise enjoy.
So that’s Warren’s business, but what’s really impressive about Warren is the other half of his professional involvement: working through the GSA [Global Semiconductor Alliance] to enhance the well-being of all players in the IP industry, not just his customers. Warren chairs GSA’s Working Group on IP, and leads the Leadership Group subset within that Working Group.
Warren also founded and continues to lead Constellations, a consortium-like group of IP vendors who meet regularly to discuss business issues, develop joint solutions, and host invitation-only events for their customers. The next Constellations event is coming up in early October.
Clearly, Warren Savage is a revolutionary, someone who believes a rising tide raises all boats in the IP industry and acts vigorously on that belief. Warren and I spoke by phone on August 22nd.
Warren Savage on All things IP …
Q: What kind of grade would you give the IP industry today?
Warren Savage: This is not something you could assign a letter grade for maturity. However, if you ask how the IP industry is serving the semiconductor market, I would give it a good solid ‘B’.
Q: Why just a ‘B’?
Warren Savage: The economics for IP companies is challenging, because there’s a lot of cost pressure pushed down through the supply chain to the IP guys. Typically, they’re the last to get involved in a project and especially for the smaller guys, tend to be surviving always on the fumes of the left-over budget of the project.
I think the smaller IP guys could do more, but they don’t have the purchasing relationship with potential customers that some of the bigger IP suppliers have.
Q: If there are only ‘fumes’ left for IP on a project, the money always goes to the major suppliers?
Warren Savage: Yes, it’s not like a shopping bazaar where there are lot of different players. Instead, there’s a trend today towards IP supplier consolidation, which leads to bigger deals between the customers and the larger IP companies while the smaller IP companies are only focusing on more boutique and specialized IP.
Q: What are the pros and cons of that situation?
Warren Savage: I see more cons than pros. The pro for the semiconductor industry is a consolidated purchasing function that makes it easier for large IP vendors to get more business and more solid traction with their customers. We see it all the time: the customer loves the IP product portfolio and loves working with the vendor. The smaller vendor, meanwhile, has even more trouble competing.
This many be good for the semiconductor company, but it completely reduces the diversity of vendors. This is the ‘con’ side of the equation, you don’t always get the best-in-breed technologies, which means the IP supply management is working to reduce the number of choices available for people who use IP.
Ultimately, of course, this phenomenon does nothing to foster true innovation in IP.
Q: This sounds like the type of complaint that’s being leveled against the EDA industry.
Warren Savage: Yes, after having spent many years myself working in EDA, the situation in IP certainly feels like it’s moving in that direction.
Q: How do customers make sure a rich tapestry of choices in IP remain available?
Warren Savage: I don’t have a simple prescription for, but it involves the customer picking suppliers they are comfortable with and then giving them as much business as they can.
Q: And doesn’t it also mean less diversity in the chips that are built on that IP?
Warren Savage: Yes, it leads to a commoditization in the semiconductor industry. A USB is always a USB, where the important thing is that the chip connects to a USB. There’s not a lot of differentiation in that type of IP. Instead, the differentiation is in how the thing is put together and, increasingly, in how the embedded software operates within that device.
Q: This potential lack of differentiation, where devices are all built with the same IP, might also push the idea of bringing design consultants on board to guarantee unique blocks within the design?
Warren Savage: Not really. If it’s a technology that’s readily available on the market, I don’t see any companies opting to go in that direction. And, these company have had a lot of experience working with consultants.
They know that if they decide to make a particular technology rather than license it, it requires that somebody within the semiconductor company is going to have to be hired to be responsible for maintain the IP code associated with that technology forever, an expense and commitment that is prohibitive. Alternatively, they can buy IP from a vendor who provides the ongoing maintenance and upgrades to the block – a more attractive prospect from the semiconductor company’s point of view.
Q: Then ultimately, IP use is a question of human resources rather than technology?
Warren Savage: Yes, because it plays into the efficiency of IP reuse. In fact, if you look at the American inclination towards efficiency, we’ve seen a tremendous move in that direction over the last few years of economic difficulty. Companies that suffered through the downturn find they can now come back to where they were before, but with fewer people, through the efficiencies of IP reuse.
When we started IPextreme in 2004, sometimes we saw semiconductor companies developing IP because there was no one at the time selling that particular type of technology. That has now changed.
At IPextreme, our idea was to utilize the output of that team, monetizing their design block beyond its immediate use and therefore returning additional revenue to the development team. All semiconductor companies are under tremendous cost pressures. Any time a cost center can be changed into a revenue center it’s a way of paying for extra engineers which leads to more innovation.
Additionally, when new standards hit the market – even before there’s IP that meets those standards – and your products need to address those standards, you have to put together a team quickly to build the necessary IP. Therefore, there’s a strategic motivation to spread your technology more widely that you can just by including it in your own semiconductor product.
Maybe you’re trying to push a new standard into the market, but you don’t want to wait for the entire semiconductor industry to wake up to the idea, read the standard, write the code, and use that code. You want people to use the standard quickly, so with a channel like IPextreme, you can speed up the process.
Q: Does that position IPextreme as a competitor to established IP vendors?
Warren Savage: Yes, that’s true. We compete against the standard semiconductor IP suppliers – we have PowerPCs and ColdFire microprocessors from Freescale, automotive microprocessors from Infineon, etc., that compete head-to-head with IP from companies like Synopsys, MIPS, and ARM.
Q: And that is a good thing?
Warren Savage: It’s absolutely a good thing for the market. It gives more choices to the customers. And the best thing about our model is that we do not sell any IP that has not already been in production with our big semiconductor customers. People buy IP that’s been licensed through us that they know has already been out there in up to a half-billion units – something that’s very comforting.
We see ourselves in that tweener-place where there’s obviously going to be a new Intel standard, for instance, and more industry-specific standards like the Compact JTAG. We can get technology that meets those standards to market faster, technology that’s going to grab hold faster – an absolutely plus for our customers.
Q: Which brings up my perpetual question about developing an organization that will benchmark IP? Will it ever happen?
Warren Savage: No, that’s not ever going to happen. The very complexity of IP does not lend itself to 3rd-party certification.
Q: Is that why VSIA evolved out of existence some years ago?
Warren Savage: VSIA was a good organization, but as the industry became more mature it naturally evolved away. It was sad in a way because VSIA did capture a lot of industry knowledge about how do thing. But as those things got freely pushed back into the industry, there was no longer a need.
For instance, when I was working in IP at Synopsys we worked hard on the Reuse Methodology Manual, among other things, and all of that effort made IP more available across the industry. And it made it possible for organizations like VSIA to fade into history.
Q: Clearly EDAC and conferences like DAC make it possible for the EDA industry to get together. Are there similar things for the IP industry?
Warren Savage: For the last few years, there are a couple of things I’ve been trying to do to that end.
First, 3 years ago this month, GSA asked me to come in and chair their IP Working Group. We’ve done a lot of interesting things so far – an ROI calculator that helps companies compare the Make vs. Buy question.
Also, we’ve created a very significant thing with an IP Licensing Survey around what are the standard business terms currently in use. It’s tricky to develop such a survey that people will participate in without worrying about exposing details of their business model.
But if both the buyer and the seller fill it out, you can use it in negotiations down the road: “Look at the kinds of indemnification IP terms that are common, and therefore this is what we should be agreeing to.”
Previous to the survey, I would talk to people and see that they had now idea of how much these things should cost. Now that we have real data, however, it’s been an extremely useful outcome of our efforts that people are negotiating from a more knowledgeable position. This it the kind of thing that is a very useful outcome of working under the auspices of the GSA. It was our epiphany that the GSA is good as surveys, so let’s put the power of their core competency together with the questions the industry needs answered.
Also, we have a Leadership Group within the GSA IP Working Group. It consists of about 10 of us – a combination of semiconductor people and IP companies. We sit down once every 6-to-8 weeks and talk about what we are going to do to further advance the state of IP. Then we use our brain trust to put together things like the survey I mentioned.
In addition, we are starting a quarterly educational seminar, an online webcast that will be available to everyone. We have many topics to cover and would like to host as many as possible going forward.
Q: So that’s your work with the GSA, but you also have other things going on.
Warren Savage: Yes, the other thing is our Constellations program. It’s kind of a collective, for lack of a better word.
Our concept is, let’s get together as a group of like-minded IP companies and form a co-op. We are working together as a team, putting some technical solutions together for our common customers – including our upcoming event in October. It’s a private, by-invitation-only event that’s a little more on the commercial side, but also trying to stay education in nature. It’s a user conference of sorts where customers and suppliers can meet together and talk through some of their mutual problems.
Also, in the spirit of what Silicon Valley used to be like, we’re going to have some fun in October as well. The event is going to include some microbrewery demonstrations, talks on beer along with some nice food – and also some wine. People who come are going to have a very good time – with the whole event being funded by IP vendors who are part of our Constellations Program.
Q: Is this a first-time event for your group?
Warren Savage: No, we had one several years ago but everyone has been too busy to get around to doing it again. Now that we’re back up and running, we expect to do this type of thing at least twice a year. In fact, we might try to do something similar at DAC next year in Austin, so stay tuned!
Q: How big is the IP industry right now?
Warren Savage: If you take out all of the patent licensing stuff, and just look at companies like ARM, Synopsys, MIPS, and IPextreme, it’s about a $2 billion business – about half the size of EDA, but growing at a much faster clip. On the patent licensing side, it’s probably in the $4 billion range.
Q: Why distinguish between the two sides of the business?
Warren Savage: Because the patent licensing, patent sharing thing is done for legal reasons and is not so much about building products.
Q: Do companies like ARM so thoroughly dominate the industry that, in fact, there’s little room for anybody else?
Warren Savage: ARM is big, but they’ve successfully proven the IP model and over the last decade have proven what the IP model can do. We hold ARM up to everyone as an example of the potential for the IP industry, even though we compete with them on some things. We sell IP into ARM customers all the time. Currently, their numbers are around $700 million, which no doubt makes them the 800 pound gorilla – but ARM has been a very positive force in the IP world.
Q: Is there still opportunity for people who want to break into the IP industry?
Warren Savage: Generally speaking, I think so. My answer is pretty tentative, because the semiconductor industry today is a difficult space for attracting investors. It’s hard for smaller companies to get a budget and compete. They first have to identify a market that they can reasonably win in, and then they have to build product and get customers – even without venture funding.
This is why today there are just about zero new semiconductor companies. That’s why I’m also involved in another GSA project called Capital Lite – it’s all about finding a way to do something about the lack of new semiconductor companies being formed in a shrinking market. Over the last decade the smaller companies have been rolled into the larger companies through acquisition, but as that happens there are fewer choices available to the customers.
Q: Are there any industry problems at this point that you’re not personally taking on?
Warren Savage: [laughing] Yeah, I’m pretty busy!
Tags: 3rd-party IP, ARM, ColdFire, Compact JTAG, Constellations, DAC, EDAC, Freescale, Global Semiconductor Alliance, GSA, GSA IP Working Group, Infineon, Intel, IPextreme, MIPS, PowerPC, Reuse Methodology Manual, Silicon IP, Synopsys, VSIA, Warren Savage
One Response to “Warren Savage: the revolutionary in our midst”