WTRS Executive Interview
Interview with Eric Broockman of CEO of Alereon .
Oct. 9, 2006
George: Can you
tell us a little bit about your background?
Eric: I've been in the industry for about 27 years. I began my career at IBM as a design engineer, rising through the ranks and ultimately becoming a senior technical staff member. Then I went to the “dark side” and went into management as a general manager. The last General Manager's position I held was in charge of the signal processing activities at IBM Microelectronics. Then I moved to Austin to work for Crystal Semiconductor in January 1995, and was responsible for a lot of the analog mixed-signal products. They had a lot of networking products, things that worked in set top boxes, and got them into the audio surround business. In 2000, I did my first startup as the CEO of Alchemy Semiconductor, which was a low-power MIPS-based processor company that was sold to AMD, and recently got sold to Raza Foundries. I then went to Legerity, and was one of the three executives there, as the VP of marketing. I also managed one of the operating divisions. I was then hired to be the CEO of Alereon, and was employee number one, starting on June 23, 2003.
That is sort of a high-level background. I come from engineering and spent a lot of time with customers from a marketing standpoint as well. I have a lot of operations background too, having been General Manager of quite of a number of different businesses over the years.
George: As CEO, do you tend to focus more on operational issues?
Eric: I would say that a great deal of my time is focused on winning and retaining customers, so that the company can scale appropriately. And I certainly have a lot of operational experience; I spend a lot of my time internally preparing for strong revenue growth as well.
George: What can you tell us about Alereon?
Eric: Alereon began as a legal entity in April of 2003. We started the company by hiring a core team of engineers from another start-up company in Huntsville, Alabama, and then moved them to Austin. Since then we’ve added quite a number of people to the company. Alereon was one of the founding companies of WiMedia, one of the founding companies for the Multi-band Alliance that became the Multi-band OFDM Alliance, and one of the founding members of the Wireless USB Working Group, which turned into Wireless USB. So, we have been at the forefront of this certification and standardization effort since the very beginning. The original team that helped start the company with me built the world’s very first Ultrawideband, FCC- certified, pulse-based radio. Then we continued that history of “firsts” as Alereon. We focused on getting chips that customers could use as opposed to building demonstration units. We have always been a chips-driven company and not a demo-driven company. In October of 2004, we were the first company to demonstrate a chip at data rates from 53 to 480 Mbps per second. And then in October of 2005 we were the first company to have a fully- WiMedia compatible chip, that was because the standards had continued to evolve, and we support all the modes and all the data rates. We recently announced at the Wireless USB Developer's Conference that our chipset has gone through full production qualification: temperature and voltage etc, and is in stock and ready for delivery and for volume production. We also continued our history of firsts there by demonstrating the world's first wireless USB device MAC. This is in support of our target market, which is primarily portable electronics. Certainly you can use a chip that is designed for portable electronics in other applications, but we decided to specialize in that regard. We have also demonstrated a sustained data rate of 480 Mbps over the air with a sustained throughput with our new MAC chip of 250 Mbps per second. So this was the first demonstration of silicon using wireless USB protocol achieving really good data rates.
George: We saw that demo last week, and one of the things that impressed us was that the demo was operating at a distance over one meter.
Eric: Yes, the team that was there told me that they ran the demo at anywhere between 2 to 4 meters, depending on what they were doing. Obviously at the lower data rates, the distance extends quite substantially. To enable our vision in the portable electronics and mobile space, which is driven by the strong value proposition UWB has in power efficiency and in moving megabytes per milliwatt, we knew that we had to get a PC host enabled, as well as other hosts, so that is why we concentrated on a PHY chip first. We chose to go on a strategic path of not doing a MAC chip for a host on a PC or on a set top box, because we had a feeling that the large semiconductor companies were going to target that. We wanted to be their PHY partner, instead of a MAC competitor of theirs. To my knowledge, all of my competitors chose to manage their risk and not trust that the really big companies were going to follow through with a MAC. They therefore split their resources and have done a host MAC and device drivers for the host, as well as for the device side. That means that their device side MAC is somewhat compromised and that their resources are diluted because they are competing with the likes of NEC, or any other company that goes into a host side MAC.
George: Is there anything else about your technology that you consider to be unique?
Eric: From a differentiation standpoint, we try to be first with production silicon that customers can ship as opposed to being early with demonstrations that don’t ship. We are focused on mobile applications, so we have the smallest footprint of any of the other solutions, and we have the highest performance. One of the things we were demonstrating last week, although we did not tout it at the show, is that we have implemented a function we call Wireless-on-the- Go. It makes our MAC behave sort of like a dual- role device, except instead of implementing the cost of supporting 127 different links, we've implemented the host function to support just one link, so that our MAC can talk to any device, be it another Alereon device or some product that has somebody else's chip in it. The reason that we wanted to do that is that if you have two teenagers in Shanghai on the bus, the probability that there is a PC on board shuttling data around is low. We didn't really want to wait until there was TCP/IP peer-to-peer on WiMedia -- that might take a while. We wanted the customer to be able to preserve the USB software interface they are familiar with and be able to use PTP or MTP or what-have-you. So we built this "on the go" feature so that we can act as a mini-host and talk to any device, to allow what therefore becomes a peer-to-peer transfer. So if you have a bunch of pictures on your camera-phone and you want to send them to your hard disk, video player, MP3-player, or video iPod, or however you want to look at them, that you would be able to do that using standard USB interfaces, using wireless USB protocol. Some companies choose, when they did the host function, to try to include a lot of the features of the whole protocol done on the host SOC. We chose not to do that because we are targeted at the mobile space, and most mobile SOCs end up with every real-time thread in the OS utilized and very little MIPS left over in their host ARM processor. So therefore we took the approach of offloading as many MIPS off the host SOC as possible and making it so that it needed no real-time threads to handle our interrupts. Therefore we differentiate ourselves because no matter what kind of processor they have, a company could utilize our solution and not worry about burdening their main SOC. Another thing that's a little bit different is that every chip company, be it a big company or a start-up, has to go through a “make vs. buy” decision for each piece of the technology. We decided for strategy reasons that we really needed to own all the key and important intellectual property for our chips. So, we did our own MAC, we did our own baseband, we did our own RF. (We actually have multiple RF teams). We own and control all our intellectual property so that if there is some issue with standards changes or whatever, we can control our destiny and therefore be timely with new products for our customers. In contrast, many of our competitors have purchased their intellectual property from companies in India (for their basic MAC function), or they have paid contractors to do their RF function and therefore, although they can do a demonstration, they don't really control their own destiny. We think that long term, if you really want to be a player, you really have to have your own technology. It doesn't mean that you might not buy a USB block from a Synopsis or something. But when it comes to the fundamental wireless USB MAC or the like, we think that is something that really has to be owned by a company if their objective is really to be an IPO company versus a company doing this for fun so they can be bought. As an example, we participated in an announcement with ARM some time ago, saying that we chose the ARM 968 processor as part of our MAC sub-system and that we chose it because when you look at it, it’s the most power-efficient ARM that they have in the ARM9 family that looks like a big ARM7. All my competitors have something that looks a lot more like an ARM924 type of hash-based chip; that uses 6 times more power per megahertz than what I use. They did that because it was cheaper, but not because it was better suited to the application. We are trying to build something that is well suited to the application and will scale over time.
We have been the only company that has done full public demonstrations of full DAA capability. We realize that we don't know what the regulatory requirements for DAA will ultimately look like, but we recognized long before others that this was likely to be important in the 3-5 GHz band. And in fact, we proposed the DAA concept to the regulators to begin with. We have demonstrated not only band dropping, but also detecting other signals and notching those signals. So we have done a demonstration where you can turn on a cell phone and you can see the overtones of the cell phone right in the UWB band. We detect that, and then notch out the frequencies around that. So we are on the forefront of a type of cognitive radio technology (we call it CogniPHY) that we think could be very useful as the UWB market develops. That is an example of some of our innovations in UWB.
George: Are you seeing a willingness in your customer base to pay a higher price for some of these features that are more focused on mobile solutions?
Eric: One of the things we did was put some very mobile-oriented interfaces on our chip like SDIO, and like compact flash. We did put a general-purpose parallel port on our chip as well, so that companies that wanted to use our chip for some PC peripherals in the early part of the market could do so. In the early part of the market, we will be participating in the PC and PC peripheral aftermarket sector through our PHY sales in partnership with MAC companies, and as well as with people who choose to use the parallel interface on our MAC chip.
From a cost standpoint, we know that the mobile market is just as cost-sensitive as the PC market so we have been pretty sensitive about implementing things that are cost-sensitive right from the beginning. One of the reasons that we did Wireless- on-the-Go versus a full USB host was not because we couldn't do a USB host, but because it takes more memory, and memory is cost, and you don't need a PC USB host inside an MP3 player, you just need enough of a host to be able to make the connection.
George: That leads into a question about sales channels. We talked to Synopsis and Eric Wong there and about having licensed the Alereon PHY. What other kinds of sales channels do you have going forward?
Eric: I had actually met with the CEO of Synopsis a long time ago. His guys were actively trying to decide if they were going to do a Wireless USB MAC or not. They were leaning that way since the IP for their wired USB had been pretty successful as a product. We both shared a vision of how large the Wireless USB market could be. Alereon participated closely with Synopsis so that they could develop a bunch of their different MACs. The reason we did this is that we would like to be part of driving the adoption curve to be as steep as possible. WiMedia, with our participation, was wise enough to put a standard MAC/PHY interface in the specification (much like 10/100 Ethernet has the SPI interface). So, since we were going to be in the PHY business, we wanted large companies to be able to get started by purchasing intellectual property from Synopsis, or from Wipro out of India, so that they could get into the game for a relatively modest cost and just use a third-party PHY. Because when a MAC goes in a system, in an SOC, the cost of the memory becomes free since you are using the main system memory instead of on-chip memory. And the die area for the rest of the chip is relatively small so that means the incremental cost for a system probably goes down 40% if you just have to buy a PHY versus buying a MAC plus a PHY, because the MAC has been integrated. So that was actually part of our strategy in working with companies like Synopsis. I don't actually think of Synopsis as a sales channel, which was the basis of your original question, but we have some of our own direct sales guys that manage geographies and key accounts and they work with major distributors that we have signed up, such as the brilliant group of Macnica in Japan, or Mirtech group in Korea, or Edom out of Taiwan. I am embarrassed to say that I forgot the name of one of our distributors in the Far East at the moment, but I just got back from vacation. So we have a set of representatives that we have been working with, and that we are very happy with, that are representing us on the ground in a variety of areas, along with our own direct sales folks. We do believe in industry partners, we do know that we are not as big as a company like Synopsis is, so we seek out opportunities to partner with a company like Synopsis or with Wipro or others.
George: As this market heads from adoption of technology into products, what sort of interest in terms of applications are you seeing today?
Eric: It’s been very interesting. We have seen a steady increase over the last few years in the whole concept of Wireless USB. If you will recall, since you have been following UWB for a while, if you rewind to 2002, UWB and wireless video were used together all the time. When we funded our company, when we went to our investors, we said, "Hey, you could certainly use Ultrawideband for wireless video, but that's not really the target opportunity we are going after. We want to get rid of all the USB cables, there’s already a large installed USB base. There are lots of applications that use USB. We want to allow those applications to be taken forward wirelessly. We think that's really the killer app. It will start in the PC aftermarket, go into the PC, go into mobile electronics, and then go into consumer electronics."
Our view was that for wireless video, there were a lot of other alternatives coming down the line. Everything from variations of 802.11a to variations of 802.11n to HomePlug2 to, believe it or not, new furniture. If you buy a new wall unit that can accommodate a big screen TV, you don't have to put the TV in a new part of the room and then worry about how you are going to get the signal to it anymore. So that is sort of the low-tech way to solve the signal to big screen TV problem. One of the driving reasons that people have the problem to begin with is that they buy the big screen TV and it doesn't fit in the wall unit they currently have. So their room is oriented in the way they want for watching TV, but now they have to put the TV someplace different when really they would sort of like to keep it the way they had it, and sometimes it’s as easy as "OK, so I'll just get rid of that old wall unit and get a new one." Problem solved. I don't know anybody who wants to keep moving their flat screen TV around just to try it on different walls.
So we approached the market probably different than many others in that era. We do think that all wireless televisions will end up with wireless USB in them, with Ultrawideband in them. But we don't think it’s because of wireless video necessarily. We do think it’s because everybody who has a digital camera wants to send their pictures to the most beautiful, largest screen they have in their house: their flat screen TV. They will edit them on their PC, and look at them there, but when they really want to enjoy them, why not put them right on your television? We think that being able to do that on an impulse, by just getting your digital camera nearby, and pushing a button on your TV, is going to be a very desirable application.
You asked about applications. In the early market we will participate in aftermarket applications like hubs and dongles and cards and so forth, through our PHY sales primarily. As the market goes into embedded applications, we see a great deal of interest in the digital camera marketplace and believe that that is going to be a pretty big market, as well as digital video cameras. Ultimately we see it also penetrating into the cell phones and becoming a mainstay, probably in feature phones first and then smart phones that have some sort of a camera phone feature and then over time becoming sort of a standard feature in the majority of all phones.
Our view is that companies like NEC or potentially Intel, or others, are likely to have an inside position to dominate the host PC market. We wanted to enable that, but don't necessarily think we have a natural inside track to that market, given what we know is happening out there. We wanted to participate to the extent that we could sell PHYs, and we really wanted to be staged so that we could enable the mobile market and help that happen faster and become as big as possible.
George: Have you been looking at wireless 1394 and some of the activities around the HANA Alliance?
Eric: I was the head of next generation networking protocols at IBM back in late 1980s, so I have a lot of experience with communications protocols starting in the mid 1980's going all the way through until I left IBM and went to Crystal. Then I had networking chipset responsibility at Crystal Semiconductor through 2000. So I have a pretty long history in the communications area, not just the chip area, and so when I looked at all of that, considered the PC side of it, it looked to me like wireless USB was really a nice bus protocol, that ultimately WiMedia will use TCP/IP, but that it will take more time. When it came to 1394, it is sort of a niche right now and it is primarily being used for video and given my earlier comments about video, I have just never been able to get excited about doing anything in that market. Certainly there are others that have and I do think that there is an opportunity there. I am hoping one of our partners takes their MAC and uses our PHY and we get some piece of that business, but we decided fairly early on that the wireless video, including 1394, was not suitable for us to target. And we also came to the conclusion that if you looked at the rate of penetration of USB in digital video cameras displacing 1394, it just didn't look like it was a good opportunity.
George: Is there anything else we forgot to ask?
Eric: What kind of company are you trying to build? We would like to the number one UWB chip company in the portable space. Our mantra, if you will, is that we want to transform and bring about "Life Without Wires." We want to build a company, maybe because I was an engineer that is the best place for engineers to work. Our objective is to be an IPO company. Our strategy is not being built around the question of what can we do to that makes it the easiest possible way to be acquired as soon as we can. That is not what we are trying to do. Statistically that could happen, and we are not doing anything that would prevent that, necessarily, but fundamentally what we are trying to do is the right thing by the customers and the marketplace to position ourselves to be an IPO. What we want is to be to the wireless USB market what CSR is to the Bluetooth market. Those are really the objectives we have as a company. The reason that I look at it that way is that for the first hundred years of radio, with the exception of the very first radio done by Marconi (which was a spark gap radio and which arguably was an Ultrawideband radio), all radios have been narrowband radios. So if I put on my big-picture thinking cap, I look at wireless USB and cable replacement as being only the first generation of applications that leverage Ultrawideband radios. It’s an enormous opportunity. It’s bigger than Bluetooth and Wi-Fi combined. So it’s an enormous opportunity right out of the box. If you look out into time, as spectrum policy evolves, other frequency bands will become available at different power levels, and my belief is that Ultrawideband will actually become much more prevalent than just cable replacement at today's tiny power levels for wireless USB. My belief is that it’s a fundamentally different approach to radio, that given the density of today's technology and the cost-effectiveness of UWB, that over time you are likely to see 5G cellular (which is really Ultrawideband based), not be narrowband based radios. So I look at it as a transition in technology. The impediment to UWB in the marketplace is regulatory policy, and the enormous financial momentum of established narrowband radio, so it’s not going to transform the industry, outside of cable replacement, overnight. But twenty years from now there are going to be an enormous number of Ultrawideband applications that are much broader than cable replacement. So I think about it in those terms. I try not to have my George Jetson hat on too much, we have to be focused, we have to get into the market, we need to satisfy today's customers, but I don’t look at this as the latest IEEE or ECMA standard that allows somebody to get a company started, get it public, and wait to see what happens. I look it as a fundamental shift in radio technology from narrowband radios of the past to Ultrawideband radios of the future and we want to develop not only an understanding of wireless USB, which we already understand well, but of Ultrawideband using technologies such as our CogniPHY DAA technology to transform the radio landscape so that we can five years from now be the company selling not only lots of wireless USB but starting to deliver what might be 5G or might be the follow-on to 802.11n. I would like to not think we would be a one trick pony. I think that this pony can go into a lot of other segments over time and our objective is really to become a billion dollar type of company.
Alereon is a fabless semiconductor company developing innovative Ultrawideband (UWB) wireless chipsets. Their mission is to simplify networking by removing cables, allowing effortless connections between PCs, consumer electronics, PC peripherals and mobile devices.
More information about Alereon here...
This interview ran in our Oct. 9, 2006 newsletter issue.