George Perlegos : We've seen wide acceptance of AVR, and with it we have managed to capitalise on the Flash technology we helped to pioneer. We've invested heavily in putting enough application engineers out in the field, by getting the product used in the education of students, and producing a strong range of development equipment to go with it. It has also attracted the attention of many third-party development tools and operating software makers who now offer a wide variety of support products to supplement Atmel's own support offerings. This wide array of support makes these that much more attractive. With over 40,000 designers using it, it is the most rapidly adopted MCU there has ever been, and it is continuing to grow at a rate of 10% per quarter. Atmel has managed to create a pretty balanced market dynamic, spreading its business fairly evenly over telecoms, computing, industry, security/identification, consumer and automotive. Is there any particular sector that you are still looking to augment further, or where you think there will be greater opportunities in the short/medium term? G.P.: We expect the most growth to come from the consumer and security sectors. Automotive is also very important, as is industrial control. We have a lot of advantages when it comes to the portable market, our MCU expertise being highly suited to the 10-20V range. Your position in the smart card market must give you a lot of confidence in the company's progression, with more and more regions using chip-and-pin for credit card security and so forth. What proportion of the company's revenue do you think these products can constitute when the market reaches full maturity? G.P.: We are currently number two in this sector unit-wise; Infineon has the top slot. With security in general becoming a key growth driver near-term, the US is starting to come on board with using this technology in bank cards and personal IDs. Currently, it is about 15% ofof our total business, but before the end of this decade, it could be worth $1 billion of revenue for us.Are the United States Government plans (following the passing of the Border Security & Visa Reform Act) to make smart card chips in the passports of every visitor to its shores by the end of this year workable, and are they likely to be more of an encroachment on civil liberties that a protection against terrorism? Is technology putting the tools in place for potential oppression? G.P.: I think you have to accept that there are a lot of inconveniences for regular travelers: all the waiting in line, the irritating questions, and so on. Let's face it, the authorities have this information anyway, if they need it they will get it. This will relieve a lot the interrogation and abuse that people have to put up with. This is technically a very challenging one to put in place in such a short time frame, so one would not be surprised if its completion were moved back slightly. Both your fingerprint sensor and smart card chips sales will be greatly increased if and when m-Commerce actually takes off, but this still seems to be an area that is making little headway. Why do you think that is, how can the situation be changed, and where will you concentrate this part of your business in the meantime? G.P.: I still believe this will take off, and we have a great future in this area. For a while, it was a question of whether the OEM thought it was too expensive to integrate this into their products. So far it has made little headway, as there is not a well-defined specification that will allow interoperable implementations, and there seem to still be some issues of who will be controlling it and who will be liable for what. Will it be laptops and the like that drive the fingerprint sensor market for the moment? Currently it is still a very small fraction of them that have biometrics built in. Is it fear of malfunction, and the inconvenience this could bring, that is holding this back? G.P.: IBM did a very long study to see how feasible it was, and so it took a while to begin to reach the market. Again, it's a convenience thing: there are new regulations in the US that expect PCs to lock within just a few minutes of being left unused. So, making the whole thing workable requires this sort of function, not to mention the fact that portable equipment now carries a lot more sensitive information, and it is easily lost or stolen. It is well documented that you are a great believer in family values, and that this perhaps stems from your Mediterranean origins. How do you feel this has affected your management technique; do you think it has been a plus point here? G.P.: I think that it is an advantage, it's all about respecting the values of other people. The company relies on the support of its employees, and if you can appreciate the contribution that the members of your team make, then everybody gains. In my office, I have a plaque that reads, "There is no limit to what can be accomplished if it does not matter who gets the credit." Everything is for the betterment of the family to the company. I feel like this makes for a nice place to work and also a place that enables our employees to be the most effective. Though you have managed to rectify the situation and are now back in good financial shape, it's no secret that Atmel was very badly hit during the downturn. Did you succeed in executing economies during that stage without losing any of your engineering prowess, or will the cuts you were forced to make become evident in the products developed in these wilderness years, which are now about to come to the market? G.P.: Most of our financial problems stemmed from too-low product margins, driven by excess manufacturing capacity insufficiently utilised. It was a tough few years, but we stuck to our principles. We put a lot of investment into product development during this period so as to make sure we could differentiate ourselves in the future. Our shareholders were not so happy at the time, but they will see the benefits in the long term, thanks to the money we put into repositioning ourselves in new sectors in which the profits will be good for many years to come. You can't go for the short-term strategy. Revenue is up by over 20% over the last 12 months. You are one of the major contributors to the Trusted Computing Group (along with companies like AMD, HP, SunMicrosystems, and IBM). This movement is clearly seeing some uptake in the commercial arena, but can it do the same in the consumer space? Is the investment needed easy to justify, or will it be a chicken-and-egg situation: you need enough people to use it to bring the cost of the silicon and software down to a price where the masses will want it, but while it remains of niche appeal, costs will stay high? Or do you expect widespread utilisation to start taking hold soon? G.P.: Yes, it will see a clear uptake in the consumer space. A couple things will drive this - the introduction of Longhorn next year and the increased concern across the board with security (for things like e-mails, web passwords, etc). But, the cost (to the consumer) will be quite low - the specification was architected to require a modest amount of silicon (but one that can include very high security). As Intel and AMD roll out microprocessors and chipsets that support the TPM and Longhorn over the next 9 months, we believe it will become increasingly common to see a TPM in a PC. Already, IBM includes it in virtually all its own PCs. Vital Statistics: Atmel Headquarters: San Jose, CA, USA Established: 1984 CEO: George Perlegos Employees: 7,194 Revenue: $1.65 billion (CY 2004)