Monday, April 5, 2010

Beceem's IPO Plans: Positive for Clearwire

On Friday after the close of market, Beceem filed its S-1, paving the way for its IPO in the coming months. The investors are mostly venture capitalists and some management participation, with Intel being the strategic investor at 20.3% (page 102 of the prospectus). In brief, Beceem is the leading play on WiMax chips, with WiMax being a broadband wireless technology deployed in over 530 networks world-wide to date.

Why is the Beceem IPO significant, and for whom? The Beceem S-1 filing comes almost to the day three years after the only other WiMax-related IPO, Clearwire. Clearwire is the dominant WiMax carrier in the U.S., with service launched in 28 markets to date, such as Chicago, Seattle, Philadelphia, Atlanta, Dallas, Houston and Las Vegas. These 28 markets cover over 33 million homes, which will more than triple to over 100 million homes by December 2010, including cities such as New York, San Francisco, Los Angles and Miami. These launches have all been disclosed in Clearwire press releases.

Here is the correlation: Every single subscriber device ("modem" for short) which Clearwire has been selling to date is based on a Beceem chip. This could be a simple USB modem or a sophisticated WiMax-to-WiFi battery-driven and pocketable converter such as the Sierra Wireless Overdrive. Of course, there are some Clearwire users who don't buy a modem from Clearwire because their laptops already come with a built-in WiMax modem, and those are made by Intel.

What is the correlation "from the other side" - Beceem's revenue going into the Clearwire network? First of all, Beceem sells to equipment makers such as Motorola and a variety of mostly Asian OEMs/ODMs, who in turn sell to Clearwire. Here is the critical sentence in the Beceem S-1 filing, from page 38: "As of December 31, 2009, over 50% of our total revenue was derived from end customers based in the United States."

Well, "over 50%" is how much precisely? Somewhere between 51% and 100%. And what part of this goes to Clearwire versus other WiMax operators in the US? I don't think anyone would disagree about the notion that Clearwire constitutes probably 99% of all WiMax investment in the US at this point. My estimate is that approximately 75% of Beceem's revenue is for deployment in the Clearwire network, splitting the difference between 51% and 100%.

So how is Beceem doing, revenue-wise? 2009 revenue more than tripled over 2008, going from $13.9 million to $43.7 million. This suggests that Clearwire is being very confident in its ability to drive very aggressive subscriber growth in 2010, as these devices produced in 2009 are going into the hands of new subscribers signing up in 2010.

Yes, I know there are numerous caveats here: Clearwire buying modems doesn't mean it will automatically gain that number of subscribers. There is inventory build. There will be other suppliers sooner rather than later. The mix with Intel-based WiMax chips in laptops is unknown. And so forth. But despite all of these caveats, these extremely high growth numbers for Beceem should be a positive sign that something right is going on at Clearwire. The confidence that Intel and the other shareholders have in filing this S-1 suggests that the pipeline for revenue growth in 2010 looks very good.

Here is an example of how the Beceem chip, operating on the Clearwire network, improves the experience for all the new Apple iPad users: Instead of ordering your iPad with an embedded HSPA modem operating on the AT&T network, just get the WiFi version of the iPad and connect it via WiFi to the Sierra Wireless Overdrive 3G/4G mobile hotspot device, powered by Beceem. This means that your iPad download and upload speeds will most likely more than double, while cutting latency (the initial lag after pressing a button, clicking a link, etc) by 70% or more. This should be a very compelling argument for any iPad, laptop or smartphone user.

In fact, this Sierra Wireless Overdrive device, operating on the Clearwire network, simultaneously connects up to five devices, sharing the $60 per month subscription. So you could argue that while delivering performance dramatically superior to AT&T, the cost per Beceem-driven Clearwire device could be counted as $12 per month ($60/5), or substantially lower than a $30 per month per device subscription from AT&T.

Ask yourself the simple questions: Would you like the wireless data network to perform better? Would you like higher speeds? Would you like lower latency? Would you like to get 250 gig per month of usage instead of 5 gig per month? Did you like the transition from dial-up to broadband some 10 to 15 years ago? If you answer yes to any of these questions, you will understand the significance of the Clearwire network and the Beceem WiMax chips powering most of the Clearwire users.

Thursday, April 1, 2010

Android's Advantage, For Now: WiMax

Sprint recently announced it will soon launch America's first WiMax smartphone, the HTC Evo 4G. This smartphone looks much like the HTC HD2 now offered by T-Mobile, but it operates on America's dominant WiMax network, built by Clearwire where Sprint is a 57% equity holder.

Let's establish the WiMax advantage over the current 3G networks operated by all major carriers. Clearwire's WiMax network has at least four major advantages over HSPA (AT&T and T-Mobile) and EVDO (Verizon and Sprint):

1. Speed. Upload and download speeds are higher. How much higher? Here is where it gets tricky. On paper, Clearwire is quoting numbers that are only approximately double that of the 3G networks. That's not the whole story, however. Which brings us to the next point.

2. Capacity. The real speed experience is similar to that of a freeway. What matters is not how fast ONE car can go, but how fast ALL cars can go when the freeway is near capacity. In the case of Clearwire, it has a lot more spectrum at its disposal than its competitors. Clearwire has 120 MHz on average per market. Compare this to what most operators purchased near 700 MHz some 2-3 years ago: around 20 MHz, or 1/6th. This means a 6x wider freeway for Clearwire. So you can fit a lot more subscribers - or have the subscribers do more things simultaneously. This is the bandwidth tonnage that matters in a battle.

3. TDD vs. FDD. Imagine if on a freeway you could dynamically, every fraction of a second, allocate the direction of any given lane. This would mean allocating the width of the freeway to the direction where there was more traffic. Effectively, this increases capacity materially. As a result, Clearwire's 6x advantage over the old networks may feel more like a 10x.

4. Latency. Speed of uploads and downloads is important, but latency is often even more important. You don't want that initial two-second "wait" before things start happening, after you click or press. In my personal experience as a Clearwire customer, I get latency around 70ms, compared to 140ms-280ms with my previous EVDO and HSPA connections. That's a 2x-4x advantage for WiMax, and makes it "feel" like a cable modem as opposed to a second-class wireless citizen.

All in all, until other WiMax smartphones are announced, the first one to hit the market this summer is the Android-based HTC Evo 4G. The superior performance of WiMax will enable users to do things where nothing else can fully compensate in terms of competitive gunpowder.

Sure, the next-generation iPhone from AT&T - and Verizon or other carriers - will bring all sorts of yet-unknown innovation to the table, probably in relation to software, content and services. But without WiMax, any such product - iPhone, Blackberry, Palm, Nokia/Symbian or Microsoft - will be notoriously handicapped in its ability to deliver a proper broadband experience.

You have to ask yourself the following question: Do you accept the premise that the biggest complaint about the current AT&T iPhone is its network performance? If you do, you cannot simultaneously say that the superiority of WiMax won't matter.

WiMax offered by Clearwire - and integrated into various Sprint-branded offerings, because Sprint owns some 57% of Clearwire - reigns alone until the fourth quarter of 2010, when Verizon launches approximately 20 MHz of FDD-based LTE, compared against Clearwire's up to 120 MHz of TDD-based WiMax. Verizon's initial LTE offerings will be primarily USB-based modems and portable WiFi hotspots, such as those built today by Novatel and Sierra Wireless. Some time in 2011, the first LTE-based smartphone will begin to appear, giving Sprint/Clearwire approximately a one-year lead.

We do not know how long Android's WiMax advantage will last. It is eminently possible that a Blackberry, Microsoft or Apple WiMax smartphone will appear on Sprint/Clearwire before 2010 is over, but I also cannot find any evidence for any of those to occur. Nokia and Palm appear very unlikely to launch anything-WiMax even in 2011 or even ever. To the extent that Blackberry, Microsoft and Apple don't launch a WiMax smartphone for the Sprint/Clearwire networks in the coming months, this will be an important advantage for Google with Android.

Verizon's advertisements often say "It's the network." The problem is that in this case, it's not their network. It's Clearwire's WiMax network, offered by Sprint, and with Google/Android powering the initial smartphone.