Accidental Entrepreneur: Cryptography Research President Paul Kocher (Part 6)

SM: What do your IP licensing structures look like?

PK: My favorite is to go to someone who is losing a ton of money. We only get paid if we solve their problem. Customers like that because it is a no-risk proposition for them and they clearly know what their problem is. The first area we where rolled that out was paid TV. The security budget per chip there was rather high so it made sense. The cost per chip could pass $1, although Moore’s Law is driving those prices down.

Today we can build something for a couple of cents that cost a dollar a year ago. Transistor costs are falling, and you can also pack something in a corner so that people don’t even know the impact of it, whereas before adding 30% to a chip area was a big deal.

We are rolling out similar technology for other areas where there is low cost per unit but high volume overall. The first project we did in paid TV had 50 million chips. That is a pretty good number. We probably ended up getting $0.50 a chip. For some projects now it is even lower per chip but with even higher volumes. Any time you have a product where the price you pay for it is much more than the manufacturing cost, you will have a problem with counterfeiting or attacks.

This applies to IP-based businesses. The replication cost of a DVD is a fraction of what it cost to put the material on the DVD. Having a genuine product able to cryptographically attest to its authenticity is important for the business model of the manufacturer. The economics justify a certain security budget which may vary widely between products. The long-term goal is to get cryptographic chips into these applications where that chip’s security can keep someone from counterfeiting the entire product. We have done this successfully in the paid TV market, and are now working towards ever lower price points for other markets. There is no reason that manufacturers cannot put something into their products which attests to its genuine nature. In that scenario, when Nike shoes come through customs, their authenticity can be verified and retailers can be sure they are not selling knock-offs. The key is to do this in a way in which the authentication itself cannot be copied.

SM: Are you seeing adoption by the Nikes of the world?

PK: They are not someone we are working with. We do have a handful of different products in different stages of the pipeline. We have a product which will be used to detect counterfeiting in an application. It will look at consumables that go with a consumer electronics device. The problem there is dealing with people who are counterfeiting those.

SM: Essentially you are enabling brand protection, right?

PK: Brand protection, quality assurance, etc.

SM: How big is this business for you? These are potentially huge volumes.

PK: Our goal is to get a few cents a chip, depending on the volume. At the end of the day we end up getting 10% of what our customers save. I want that to be a higher number. Still, when you start with a company that is losing hundreds of millions of dollars a year, the improvements can be significant amounts.

This segment is part 6 in the series : Accidental Entrepreneur: Cryptography Research President Paul Kocher
1 2 3 4 5 6 7