
If one believes futurists, the coming decade will be very bandwidth intensive indeed.
In the developed world, forecasters at the World Future Society envision an era of near-ubiquitous broadband. Smartphones, PCs and gadgets yet to be popularized will download high-definition video at ultra-rapid speeds from virtually anywhere. Sufferers of chronic conditions will monitor their bodies continuously through sophisticated wireless sensors that also connect to health care providers. In developing countries, citizens will have access to sophisticated financial services through their mobile phones.
Those will be the visible manifestations of progress. But neglected by most armchair futurists are the efforts required on the backend. The pipes and plumbing of the Internet and wireless networks will need a significant upgrade over the next several years, and a costly one at that. Fiber-optic networks—which had enough of a building boom a decade ago to produce an oversupply of bandwidth for years—are finally hitting capacity. Wireless networks are even worse-equipped to keep up with demand for data. A recent report from Cisco Systems projected that global mobile data traffic will increase 39-fold from 2009 to 2014.
“There’s no way you could use today’s technology to accommodate all this traffic,” says Marty Cooper, an entrepreneur and angel investor best known for leading the team that developed the first cell phone in 1973. Just as fiber-optic networks require ever-more-advanced technology to cram more data over a single strand, Cooper says, wireless networks must evolve to more efficiently serve an ever larger and more sophisticated array of mobile devices.
Where do VCs fit in? “There’s oftentimes opportunity during a generational shift,” says Jon Auerbach, general partner at Charles River Ventures, and a member of the 4G Venture Forum, a collaboration between venture firms, Alcatel-Lucent, Ericsson and Verizon Wireless, focused on identifying and commercializing 4G technologies. He compares the current period to the transition from analog to digital mobile networks in the 1990s, noting that “the shift to 4G is such a large architectural shift that it will open up the playing field.”
Granted, it won’t look anything like the great telecommunications infrastructure investment bubble of 2000, when venture capitalists invested more than $17 billion in 900-plus companies developing tools and equipment for building and managing wireless networks and broadband infrastructure, according to Thomson Reuters (publisher of VCJ). In the 12-month period ending in the first quarter, venture capitalists invested $3 billion in 224 deals in those sectors. Still, that’s an increase of about 20% from the same period a year earlier, with some sizeable deals in the mix.
In September, for example, a consortium of private equity and venture firms made a $360 million investment in NextG Networks, a 9-year-old developer of fiber-optic distributed-antenna systems. And in January, VCs invested $125 million in Extenet Systems to build out its business of providing shared communications networks for wireless service providers.
Today’s investors are particularly interested in technologies that extend the reach of broadband networks. Recently funded companies in that category include M2Z Networks, a provider of last-mile wireless broadband access, and Tower Cloud, which provides extended backhaul connecting wireless towers and switching centers. VCs are also putting a sizable chunk of their money into companies that first raised capital over a decade ago. Among the old-timers that have raised fresh capital recently are two 11-year-old companies: Actelis Networks, a developer of Ethernet-over-copper networking equipment, and IPWireless, a mobile broadband solutions provider.
Exit prospects, meanwhile, look reasonably bright for developers of technologies for next-generation communications networks. After 27 months without a single IPO from a VC-backed communications equipment company, no fewer than three such companies went public in March—Calix Networks, MaxLinear and Meru Networks—raising nearly $240 million combined (see “IPO Hat Trick” sidebar). And despite the fact that two of the three have been unprofitable for years, all three priced at the top of their range or above it and all posted double-digit percentage gains in first-day trading.
There’s oftentimes opportunity during a generational shift. The shift to 4G is such a large architectural shift that it will open up the playing field.”
Jon Auerbach
The outlook for acquisitions is equally bright. Cisco CEO John Chambers indicated in his last quarterly conference call that the company intends to be “very aggressive” in its acquisition strategy—this after spending $2.8 billion in October to purchase formerly venture-backed Starent Networks.
Cisco’s competitors are shopping for venture-backed companies, as well. Broadcom, for example, has purchased two since December: Teknovus and Dune Networks, both of which make chipsets to connect devices to high-speed broadband networks. Broadcom paid $178 million in cash for Dune, which had raised about $50 million in VC funding, and $123 million in cash for Teknovus, which had about $66 million in venture backing, according to Thomson Reuters.
Yet VCs have reason for trepidation about communications deals. For one thing, many are still holding onto telecom portfolio companies first funded more than eight years ago, with more than a few of them racking up more than $100 million each in venture funding.
But for venture-backed companies that make it to market with the right product at the right time, rewards are enormous. Just ask any of the 1980s investors who risked a few million on a much-rejected startup called Cisco.
Following, we look at how VCs are placing their bets in the big broadband gamble. From 4G smart antennas to 100-gigabit Ethernet, there’s certainly a higher level of complexity to developing technologies than can possibly be squeezed comprehensively into these few pages. The intent, therefore, isn’t so much to suss out the winner in the LTE chipset battle, but to discuss the emerging problems venture-backed startups are addressing, highlight up-and-coming players and later stage turnarounds, and attempt to identify from which corners the next multibillion-dollar exits will emerge.
THE COMING CRUNCH
First, the problems.
“Too much is never enough” was the tagline for MTV in its early days. It could just as well apply to broadband infrastructure. For years, fiber-optic network operators suffered from a glut of capacity after a massive building spree in the late 1990s. Wireless carriers, meanwhile, struggled to market higher-end data plans to consumers who saw little value beyond text messaging.
That all changed dramatically a few years ago, says Shirish Sathaye, a general partner at Matrix Partners, which has invested in several wireless infrastructure startups, including three that have exited in the past three years: Airvana, which makes wireless broadband hardware and software; Aruba Networks, which develops an enterprise network access platform; and Starent, which develops software and hardware to deliver multimedia content to mobile devices. Most recently, Sathaye led a Matrix investment in SpiderCloud, a developer of technology for bolstering cellular coverage at employer sites.
In terms of core wireline network capacity, there is enough, but it is in the wrong places, not near the cell towers.”
Shirish Sathaye
With the advent of Apple’s iPhone and other high-end smartphones, carriers quickly found their networks weren’t as robust as they thought. “In terms of core wireline network capacity, there is enough, but it is in the wrong places, not near the cell towers,” Sathaye says. Mobile carriers need faster, more reliable and more numerous connections to backhaul networks. It’s a global problem, he adds, but mostly concentrated in the United States, where more people live spread out in suburbs.
It’s an issue in urban areas, too, says Ronny Haraldsvik, vice president of marketing at SpiderCloud. “You may find yourself in San Francisco wondering, ‘Why is my connectivity so bad when I have five bars?’” he says. “It’s become a misnomer, because coverage doesn’t mean that you have capacity.” Just one-third of U.S. mobile users have smartphones, but they consume up to 70% to 80% of capacity on networks, Haraldsvik says.
Much of the wireless bandwidth problem has been created by continuously connected applications, such as Twitter, which constantly ping the network for updates. That’s a problem because networks were built principally for voice traffic. Without radical measures, Haraldsvik predicts, problems will only get worse when droves of people begin using their handsets for applications such as high-definition video, peer-to-peer, and Skype.
It’s an accelerating trend. Cisco predicts that within four years video will account for 66% of mobile traffic. That same year, it’s likely there will be more than 5 billion personal devices connecting to mobile networks, and billions more machine-to-machine nodes such as sensor networks. Exponential growth is already occurring, with global mobile data increasing 160% year-over-year in 2009, and growing 2.4 times faster than fixed broadband data traffic.
And it’s not just a wireless capacity crunch for carriers, says Jonathan Lacy, CEO of Ofidium, an Australian developer of 100 Gigabit-per-second transmission technology for fiber networks. “They have a lot of fiber in the ground, which is good,” he says. “But in many cases, that fiber is almost at its limits in terms of today’s technology.”
WORKING SMARTER
Luckily, there’s no shortage of entrepreneurs pitching at least partial solutions to the capacity crunch.
Marty Cooper, 81, is one of the more outspoken. Between keeping an eye on his two startups—wireless “smart antenna” developer Arraycomm, which he founded in 1992, and simple-to-use mobile phone provider Jitterbug, which he co-founded in 2006 with his wife, Arlene Harris—Cooper spends his time filling in legislators on how they ought to regulate spectrum. His principle talking point is on “spectral efficiency,” or policies and practices that promote the use of existing airwaves to provide ever-richer data and applications.
Carriers have already convinced Federal Communications Commission leaders to allocate a greater portion of spectrum in coming years for wireless communications applications to alleviate some capacity issues. But efficiency mavens contend it will be technology, rather than an expansion of spectrum licenses, that will enable next-generation applications like high-definition video over mobile devices.
You may find yourself in San Francisco wondering, ‘Why is my connectivity so bad when I have five bars?’ It’s become a misnomer, because coverage doesn’t mean that you have capacity.”
Ronny Haraldsvik
“There is a myth that there’s a scarcity of spectrum,” Cooper says. “The reality is it’s not the amount of spectrum. It’s the capacity of the spectrum to deliver information.” His “Cooper’s Law” states that every 30 months the amount of information that can be transmitted over the useful radio spectrum doubles.
Cooper has been pitching a partial solution to the capacity crunch—Arraycomm’s “smart” antenna” systems—for the past 18 years. But even after $122 million in venture funding, he estimates the company is a couple of years away from profitability.
Cooper says there’s still plenty of room for progress because cellular networks waste spectrum by under-using data compression technology, next-generation signal processors and other tools that would enable transmission of much clearer calls and richer digital content.
VCs are trying to correct that inefficiency. Since last year, venture firms have backed a plethora of startups that could be described as spectral efficiency plays. One of the larger rounds went to Petaluma, Calif.-based Cyan Optics, a supplier of packet optical transport platforms that raised $22 million from Azure Capital Partners, Norwest Venture Partners and Tenaya Capital.
Cyan plans to use the money to further develop a line of products that enable backhaul network operators to meet capacity needs of wireless providers. The company’s founder, Michael Hatfield, previously co-founded newly public Calix and Cerent, a provider of high-speed fiber-optics systems acquired by Cisco for $7 billion in 1999. Hatfield says the genesis for Cyan came from the observation that mobile carriers will be looking for new, lower-cost ways to dramatically scale network capacity after they see traffic surge on relatively flat revenue.
SpiderCloud was also the recipient of a large funding round, raising $25 in February from Charles River, Matrix Partners, Opus Capital and Shasta Ventures. The company, based in Santa Clara, Calif., makes a product called an Enterprise Radio Access Network (E-RAN) that can be used to manage wireless usage within a company or government office. SpiderCloud hopes to deploy highly dense radio nodes at indoor locations that can sharply and inexpensively boost providers’ ability to offer mobile broadband.
In the wireless world, in particular, says Charles River’s Auerbach, carriers’ needs have changed so dramatically in the past two years that it’s created a sweet spot for “very disruptive technologies” to help reduce costs and increase revenue. Though the major carriers haven’t historically been big acquirers of venture-backed companies, they are often the earliest and most crucial customers. That was the case, for example, with Starent, a onetime Charles River portfolio company that Auerbach says got its first big marketing win by signing up Verizon as a customer.
HEY, BIG SPENDER
Market researchers also see startup gear makers in a comparatively strong position for securing sales. Over the next five years, analysts predict a marked jump in spending on both fixed and wireless broadband infrastructure and services.
There is a myth that there’s a scarcity of spectrum. The reality is it’s not the amount of spectrum. It’s the capacity of the spectrum to deliver information.”
Marty Cooper
Wireless is expected to see the most dramatic jump. In 2013, service providers are projected to collect more than $877 billion from cellular customers, up from an estimated $624 billion in 2008, according to research firm Infonetics. They’ll also be spending on 4G upgrades, split between LTE and WiMax, the two predominant 4G technologies. (WiMax has been adopted by Sprint and many carriers in developing countries, while LTE, or Long Term Evolution, is backed by Verizon and AT&T.)
Infonetics forecasts the market for WiMax equipment and devices will hit $4.97 billion in 2013, while the market for LTE infrastructure, largely perceived as winning the standards war in developed economies, will top $5 billion by 2013. Even at such levels, however, it’s unclear whether upgrades will keep pace with mobile data demand. “I think the data network will handle a lot more than what consumers want now, but it is going to be extremely bandwidth constrained when all content is available in HD and available for streaming on demand,” says Satish Dharmaraj, a partner at Redpoint Ventures and a 4G Venture Forum member.
Meanwhile, spending on fixed broadband services has also been growing at double-digit rates. In 2009, global fixed broadband service revenue totaled $164 billion, a 13% increase over 2008, according to estimates from ABI Research. ABI predicts broadband service revenue will exceed $210 billion in 2014, with demand for services such as IPTV and online gaming overriding recessionary belt-tightening.
Supplying all that bandwidth won’t be cheap. Between 2009 and 2013, Infonetics forecasts service providers worldwide will spend a cumulative $146 billion on carrier Ethernet equipment, with a focus on the fastest standards: 10 GbE, 40 GbE and 100 GbE.
While the largest suppliers of routers, switches and optical gear are prominent public companies such as Alcatel-Lucent, Cisco and Juniper Networks, venture-backed companies are also carving out a niche. Several have raised funding in the last couple of years. Fremont, Calif.-based Actelis Networks, which develops technology for delivering Ethernet over copper networks, raised $15 million last year from a dozen firms, including Adams Street Partners, Carlyle Venture Partners, Global Catalyst Partners and Walden International. Actelis believes only a limited number of locations that require fiber actually have access to it, and that copper networks can be equipped to deliver bandwidth for mobile and fixed broadband.
Morrisville, North Carolina-based Overture Networks, a developer of products for delivering Ethernet-based services over copper and fiber, raised $18 million in January 2009 from Intel Capital, Intersouth Partners and Morgenthaler Partners. About one month earlier, 10-year-old Overture, which has raised $63 million to date, acquired venture-backed Ceterus Networks, which develops optical Ethernet access technology for cell site backhaul.
Early stage funders are active, too. Australia’s Starfish Ventures invested $4 million in Melbourne-based Ofidium last year from to further develop so-called orthogonal frequency division multiplexing technology it says will reduce impairment of data over long transmission distances. CEO Lacey says he believes there are more startups active in the space than a simple Web search would uncover, as many are in stealth mode or haven’t raised venture funding yet. “If it weren’t for the financial crisis, we would probably be hearing about more optical communications startups,” he says, adding that he expects more will emerge from stealth in the next year.
ROOM FOR IMPROVEMENT
Despite a several slow years for liquidity events, industry watchers say the climate is looking brighter for developers of applications and gear for broadband networks. However, that needs to be taken in context.
The data network will handle a lot more than what consumers want now, but it is going to be extremely bandwidth constrained when all content is available in HD and available for streaming on demand.”
Satish Dharmaraj
“It just hasn’t been very active,” says Peter Morgan, managing director at The Spartan Group, a boutique investment bank with offices in Los Angeles and San Francisco. Morgan says he recently put together a chart of global IPO activity in 2009, ranked by sector. Telecommunications and telecom services came in dead last, with just five offerings.
But the IPO market sector mix is changing this year, according to Renaissance Capital, which provides IPO research and investment management services. In its Q1 overview, the firm said it expects to see a continued shift toward traditional growth sectors within the technology industry.
On the heels of March IPOs from Calix Networks, MaxLinear and Meru Networks, three more communications-related companies hope to follow in the trio’s slipstream:
• Aeroflex Holding Corp., which makes chips and microwave and RF devices for the aerospace, defense, fixed broadband, wireless/mobile, and test and measurement markets, filed for a $500 million IPO on the New York Stock Exchange in April. The Plainview, N.Y.-based company’s customers include Boeing, Cisco and Raytheon. Aeroflex was taken private by Golden Gate Capital, Goldman, Sachs & Co. and Veritas Capital in a leveraged buyout in 2007, according to Thomson Reuters. Veritas is its largest shareholder (with a 40.34% pre-IPO stake), followed by Golden Gate Capital (23.45%) and Goldman (19.7%), according to the company’s prospectus. Aeroflex lost about $77 million on sales of $599.3 million for the year ended June 30, 2009.
• Beceem Communications, a wireless chipmaker, filed in April for a $100 million public offering on the Nasdaq. Two months ago, the company introduced a chip for 4G WiMax networks that employs decoder technology and interference-cancellation algorithms to increase the performance of mobile devices at both high- and low-signal strength. The 7-year-old company has raised $130 million in venture backing to date. Intel Capital is its largest shareholder (with a 20.3% pre-IPO stake), followed by Global Catalyst Partners and Walden International (each with a 16.3% pre-IPO stake) and Khosla Ventures (5.9%), according to Beceem’s S-1 registration statement.
• And Force10, an Ethernet products and services provider, hopes to raise up to $144 million through an NYSE offering. The San Jose-based company previously raised $418 million in venture funding, according to Thomson Reuters. In 2009, it acquired rival Turin Networks, an 11-year-old company that had raised $254 million in VC. The largest current venture stakeholders in Force10 are Advanced Equities Capital Partners (28% stake) and New Enterprise Associates (9.8%).
REASON TO WORRY
Despite the success of recent IPOs, VCs have cause for wariness. In a startup environment in which entrepreneurs in fields such as social networking and software-as-a-service routinely get from concept to product launch on a personal credit card, next-gen networking equipment companies can be remarkably costly. Force10, for example, posted a loss of $55 million in its last fiscal year—more than 10 years after its founding.
Moreover, carriers have a history of taking longer than expected to execute on planned upgrades and of dropping strategies that prove too costly. For example Verizon, spent close to $23 billion on the initial phase of its Fios fiber-to-the-home network build-out, but it is now winding down the effort. On the wireless side, carriers also have an incentive to squeeze out maximum value from existing networks. “The well-established cellular operators have spent billions of dollars deploying 3G networks, so they’ll be hoping to see a return on their investment before committing too heavily to 4G rollouts,” says Peter Bell, an analyst at research firm Telegeography.
The well-established cellular operators have spent billions of dollars deploying 3G networks, so they’ll be hoping to see a return on their investment before committing too heavily to 4G rollouts.”
Peter Bell
Then there’s the risk of being too late. If one listens to futurists, it’s easy to imagine that even the most cutting edge of 4G companies won’t make it in the long run. The World Future Society predicts that within a few decades we may not even be using airwaves to transmit data. Instead, we’ll get speeds at least 100 times faster using lasers.
IPO Hat Trick
Prior to March, the last VC-backed communications company to go public was Neutral Tandem, which made it out in November 2007.
But the 27-month drought ended in March, with IPOs from Calix Networks, MaxLinear and Meru Networks raising close to $240 million combined. All three priced at the top of their range or above it and all posted double-digit percentage gains in first-day trading.
Calix Networks
Calix, a Petaluma, Calif.-based provider of communications access systems and software, priced its 6.3 million-share offering at $13 per share, the high end of its proposed range. The stock closed up 16% in first-day trading on March 24, providing an initial market cap around $472 million. Wall Street didn’t mind that the company has posted annual losses for the past five years, including a $22.4 million loss on revenue of nearly $233 million last year.
Prior to going public, Calix raised about $220 million in VC, according to Thomson Reuters (publisher of VCJ). Its largest VC shareholders were Foundation Capital (9.06% pre-IPO stake), TeleSoft Partners (8.37%), Azure Capital Partners (8.23%), Meritech Capital Partners (6.59%) and Redpoint Ventures (6%), according to the company’s prospectus. Other venture backers, including Integral Capital Partners and Menlo Ventures, held pre-IPO stakes of less than 1% each.
MaxLinear
Like Calix, broadband chipmaker MaxLinear was given a warm welcome by the Street. The Carlsbad, Calif.-based broadband chipmaker raised $90 million in its IPO, which priced slightly above the anticipated range and gained another 18% in first-day trading. Previously, MaxLinear raised around $35 million in VC funding, according to Thomson Reuters.
Its largest VC shareholders are U.S. Venture Partners (21.62% pre-IPO stake), Battery Ventures (13.75%), Mission Ventures (13.03%) and UMC Capital (7.09%), according to its prospectus. MaxLinear earned $4.33 million in net income on just over $51 million in revenue last year.
Meru Networks
Meru made it a hat trick on the last day of March. The maker of wireless LAN equipment priced 4.4 million shares at $15 apiece (the top of its range), raising nearly $66 million. Traders weren’t put off by the fact that Meru posted a net loss of $17.4 million on revenue of $69.5 million last year. The company’s stock opened at $20 and closed at $19.17 on its first day of trading, up 28 percent.
That was good news for the Sunnyvale, Calif.-based company’s venture backers, who put $171 million into Meru since 2002. Its principle shareholders include Clearstone Venture Partners (21% pre-IPO stake), Vision Opportunity Master Fund (19%), NeoCarta Ventures (18%), Bluestream Ventures (16%) and D.E. Shaw (16%). —Joanna Glasner