Trends in the Wi-Fi industry - Michael Hurlson of Broadcom

Just before Broadcom spoke at the Computex CE Summit on June 4, Digitimes had the opportunity to chat with Michael Hurlson, vice president and general manager for the Home and Wireless Networking Business Unit at Broadcom about trends and topics in the global WLAN industry.

Q: What trends in the Wi-Fi industry are being discussed at Computex Taipei 2008?

A: I think this year the big issues we are seeing in the market, or perhaps we should call them opportunities, are the opportunities to put Wi-Fi into a host of new products.

If you look back to one year ago at Computex, I think the challenge at that time was how to stimulate the 802.11n market. There had been some downstream confusion about the status of 802.11n early in 2007, but after Computex in June 2007, the Wi-Fi Alliance introduced a testing and certification program for 802.11n products and the market started picking up through the second half of 2007 and early this year. So, if you look back now and examine how the market reacted to the challenges we faced last year, I think our customers would largely say that it was mission accomplished.

However, much of the development over the past year was focused on the traditional markets of players in the WLAN industry – PC and home router applications. And although there continues to be interesting opportunities in these markets, they are also extremely competitive and face continued price pressure.

So going back to my original statement, I think the opportunities for the Wi-Fi industry this year are to see how we can expand our markets and bring additional revenue opportunities by building a new family of products that incorporate Wi-Fi.

More specifically, we see two opportunities starting in Computex in 2008 and running through 2009. One is around mobility, meaning incorporating Wi-Fi into a host of new handheld products such as handsets, MP3 players or portable gaming devices. The other is building a set of products and technology backbone that enables us to move video off a broadband network, bring it into the home and move the video from the gateway over a wireless link to an IP set top box (STB) for viewing on a TV.

Q: Wi-Fi has advanced to 802.11n, so where can we expect Wi-Fi to advance after that?

If you talk about evolution in the market, sometimes there is too much focus on technology. For years in the WLAN market the discussion has been about when are we moving from 802.11b to 802.11g, or from 802.11g to 802.11n. And now the question is what comes after that.

For Broadcom, we focus on this question from an applications perspective rather than from a technology prospective. At this point in the market's evolution, there is more than enough bandwidth for applications using Wi-Fi. But the question remains as to whether there are enough applications using Wi-Fi.

So the challenge for Broadcom was how to create opportunities for integrating Wi-Fi into more and more systems, and on the handheld side that means handsets, MP3 and portable gaming applications.

By monitoring the handheld space, we noticed that currently the Wi-Fi intersection is 1-1 with Bluetooth, meaning that if a product has Wi-Fi, it will also ship with Bluetooth. Unfortunately, that also means that Wi-Fi would normally be implemented into multiple radio solutions, and when vendors start offering multiple radio solutions or multiple connectivity options in a product, their designs face increased challenges concerning space, power and cost.

To solve this issue and create more opportunities for Wi-Fi in the market, we've designed a SoC (system on chip) that combines Wi-Fi and Bluetooth, as well as incorporating an FM receiver. We've also taken it to a very small form factor by using a 65nm manufacturing process. Looking at some other benefits, an integrated solution has a power budget that is much lower than using discrete Wi-Fi and Bluetooth, and we are offering the SoC at pretty compelling price points.

So, in one fell swoop we've addressed the concerns of space cost and power, leveraging one of Broadcom's assets, which is integration.

The product has been under development for some time. We launched the chip in early 2007 and we still haven't talked about meaningful production. But now we expect to talk about that soon and with some very specific customer names attached to it. Therefore, we are very optimistic about the opportunities this presents to the mobile market and us. For example, in the handset market, we expect Wi-Fi traction to move from low single-digit attach rates to double digits and hit 20% in 2008-2009.

Q: It wouldn't be Computex without a Digital Home discussion, which dovetails nicely with the opportunities you mentioned of bringing video into the home via broadband and then sending it over Wi-Fi to a TV. This has been discussed for years, what makes this year different?

A: Remember, we need to go back to what I said about the challenges that the market faced last year. Now that 802.11n has gained traction, the parameters in the market are completely different. Without an 802.11n infrastructure, the ability to distribute video over Wi-Fi is a big challenge. It can be done but it's inconsistent and it's not what we consider carrier class.

For the concept of streaming video in the home to have a significant impact in the market, the assumption we had was that an 802.11n gateway was needed to enable this to happen. Another assumption was that competition would drive the market in such a way that 802.11 SKUs were being offered at an affordable price. With the market scaling toward 802.11n, we now have a starting point to take full advantage of this opportunity.

So, throughout the course of 2008 and 2009, Broadcom will offer and roll out products that focus on video distribution over Wi-Fi. Broadcom's advantage in developing this arena is that we've leveraged our video heritage and put this intelligence into products. We already have strengths in the cable and satellite markets, and we have years of experience knowing how to handle video signaling. For example in our satellite STB platform, we have extensive experience dealing with and handling some of the disturbances that naturally occur in the wireless distribution of video.

So when we looked at how to handle video signals in our next generation of products, we brought some of that intelligence between the video layer and the network layer, and we've specifically leveraged how we buffer packets on the video layer with how we deal with it on the network layer. What we've done is incorporate the learning from video and implemented that into a set of WLAN products that include software and chip architecture, and these solutions will be implemented into wireless routers or wireless gateway products.

Q: Can you comment on the status of IPTV and triple play services in the market?

A: IPTV was talked about a great deal but the market hasn't taken off to a degree that anyone would like to see. However there is renewed excitement around IPTV from the carriers themselves. They are aggressively targeting the IPTV market and offering services over IPTV. We've seen it in the US and it seems to be resonating very well with the customer base there. Next-generation networks in Japan will also have IPTV offerings for customers, and in Taiwan, Chunghwa Telecom appears to have taken a much more aggressive approach with respect to how they are offering these services.

If you look at the historical issues that the market has faced, most have been resolved. First of all, there were questions about having content being ready for IPTV, but that issue seems to have been largely resolved in the market. Another problem was around the technology. But if you look at Broadcom, we have a compelling IPTV STB chipset solution that carriers find very attractive. So we have content coming in and we have hardware that is ready, and for Broadcom's part we are competing by aggressively offering low-cost solutions to our customer base. But the issue that has surfaced in the market is how quickly service providers can capture customers.

The problem is the capture rate. An inhibitor to customer capture rates on the IPTV side is that each customer that signs up needs a truck roll, meaning someone has to go to the subscribers residence and install the wires between the gateway and the TV, or multiple TVs where the service would be viewed. And that requires a day, or a half-day of labor and it's an extremely costly issue for carriers and consumers alike.

The carriers have asked Broadcom to try to help them try to resolve this issue and our advice to them was to make the process of installation much easier by letting us go offer a wireless link based on 802.11n and promote this to potential customers as an option for self install. This type of solution drives down the cost of installation for the carriers and drives up the number of customer that these carriers are able to capture on a daily basis, hence driving up their capture rates. Carriers, who are ultimately responsible for offering triple play services, have the pieces to the puzzle but they don't have the connectivity solution to enable it.

Q: Going green is a hot topic in the market. How does Broadcom deal with power issues with its products?

A: Our customers are increasingly concerned about power, and this is not only in the mobile segments. People are seeing the costs of power rise steadily amid daily price hikes in the oil industry and customers are showing a strong desire for wanting greener products, so concerns about power run across the entire gamut of applications, from Wi-Fi, to DSL modems, to DTV solutions. So for Broadcom, we've approached this issue by incorporating intelligent design into our products by better managing a down state by turning off the sub-systems that aren't being used, and intelligently managing power consumption on all products.

But in addition to design solutions that gate off or intelligently manage areas of chips that aren't being used in particular situations, we have chosen to take full advantage of the benefits advanced manufacturing processes offer concerning power consumption. One thing that Broadcom has done very well is that we have been able to scale down our manufacturing to advanced geometries and all of our new solutions are built on a 65nm process. With the smaller geometries come the associated power efficiencies that customers demand.

Q: With Intel introducing more features into its mobile platforms, does this make it difficult for Broadcom to complete in the notebook space?

A: If you look at the market for integrating Wi-Fi in notebooks, we've been able to successfully compete for a very long time. It's an important part of our business and will continue to be an important part of our business. Our products can be found in the product lineups of almost all of the top-seven notebook vendors. And while they use Broadcom Wi-Fi in some conjunction with Intel Wi-Fi in their lineups, we've actually seen our share in the notebook space increase over the last year (from Computex 2007 to Computex 2008). Our increasing share has been driven by our ability to offer attractive consumer grade platforms, which have been the fastest growing segment in the market. By consumer grade platforms I am referring to a basic Wi-Fi offering, which would be 802.11g.

While I mentioned before that one of Broadcom's strengths is our ability to compete in the integration vector, another strength is our ability to compete on cost. So these more cost sensitive platforms play to our strengths.

In addition, PC vendors have started to go into new product classes such as UMPCs and mobile Internet devices (MID), as well as ultra mobile PCs, which have received the most media attention recently. These products are designed to require only basic connectivity, typically across an 802.11g link. And as these products become increasingly popular, I think we will be increasing our share again moving through calendar year 2009.