WiMax: Hyped, but not Backed?

WiMax is among the most promising, and well-hyped, wireless networking technologies around today, but to reach even half the ambitious performance targets set for it will require a step-change in wide-area radio performance.

Key to achieving this are the capabilities of the chipsets that will form the heart of WiMax base stations and customer devices. But for a networking initiative with so much riding on it, investment in this crucial underlying technology is underwhelming, the latest Unstrung Insider report details.

According to the report, entitled WiMax Chipset 2006 Market Outlook, seven startup 802.16 chipset developers had, by January 2006, raised $170 million among them, of which $60 million was secured in 2005. This is shown in this table, excerpted from the report:

Table 1: Investment in WiMax Chipset Startups
Vendor Founded; Location Funding to Date Value & Date of Last Round
Beceem 2003; Santa Clara, Calif./
Bangalore, India
$40M $20M Series C, Dec. 2004
Cygnus 2004; Carlsbad, Calif. $20M $20M Series A, July 2004
PicoChip 2000; Bath, U.K. $41.3M $20.5M Series C, June 2005
Runcom 1997;Tel Aviv, Israel $8M $8M Series A, 1997
Sequans 2003; Paris $11M $9M Series B, Feb. 2005
TeleCIS 2000; Santa Clara, Calif. $18.7M $10M Series C, Dec. 2005
Wavesat 1993; Montreal $30M (complex history) $10M, Oct. 2004
Source: Unstrung Insider

Now, $170 million is a lot of money by any standard; but given the scale of the challenge, it may not be sufficient. Adding to that challenge is the fact that WiMax is less tightly defined than other networking technologies, such as 3G, WLAN, or DSL. Chipsets for Fixed WiMax (802.16-2004) and Mobile WiMax (802.16e), for example, require fundamentally different designs, and are developed with a wide variety of end-user devices in mind – from base station, to outdoor CPE, to desktop modem, to notebook card, to mobile terminal.

Diluting the Money
Trying to design products for all of these types of device dilutes R&D resources, with the effect that WiMax startups have typically specialized in different market segments. The effect of this, in turn, has been to dilute the impact of that $170 million aggregate investment.

The two companies that have received the greatest investment to date – Picochip ($41.4 million) and Beceem Communications Inc. ($40 million) – are focused on base stations and mobile terminals respectively, and therefore have different design requirements and product development strategies. Looked at another way: Only two out of the seven startups – Beceem and Runcom Technologies Ltd. – are focused primarily on the mobile handset market.

The aggregate $170 million outlay on WiMax silicon startups also compares unfavorably with investment in 802.11 wireless LAN. As a rough comparison, total investment in the eight wireless LAN startups featured in our July 2003 Wireless LAN Chipsets market report was $327 million – almost twice the current level of investment in WiMax chipmakers.

This relatively low investment in WiMax is somewhat surprising, given the high returns risk-taking investors have made from acquisitions and IPOs of DSL, wireless LAN, and even 3G silicon companies over the past few years.

Waiting For the Big Boys?
One explanation could simply be that venture funds are not interested in a technology that faces numerous hurdles and an uncertain deployment timetable. North American VCs, in particular, have many well-rehearsed arguments (spectrum, competition from 3G, etc.) for why WiMax won't succeed, or isn't yet ready to succeed.

Another explanation could be that investment in startups is held back by the perception that WiMax is a big-company play. With Intel Corp. (Nasdaq: INTC) in the game, what chance does a startup have?

That's a hard line of thought to argue with, but it's dangerous to build global networking businesses on a foundation of limited competition from the supplier base. Sure, the big system houses that have committed to WiMax – the likes of Alcatel (NYSE: ALA; Paris: CGEP:PA), Motorola Inc. (NYSE: MOT), Nokia Corp. (NYSE: NOK), and Siemens AG (NYSE: SI; Frankfurt: SIE) – have done so partly on the promise of the "free CPE" from Intel (i.e. WiMax embedded in a notebook, and therefore "free"). But it's anathema to their usual thinking for these big vendors to cede market control and pricing power to their suppliers.

Realizing this, strategic corporate investors have had to step up to the mark and walk where VCs fear to tread.

Corporate Money Steps In
Even Intel, which has a major market development program underway designed to support WiMax at all levels in the value chain, has invested in a not-quite-rival base station and high-end user-terminal silicon provider, PicoChip, through its Intel Capital fund. And Samsung Electronics Co. Ltd. (Korea: SEC), having similarly passed the point of no return on WiMax, has invested in Beceem and, we understand, contributed to mobile WiMax development work by Runcom.

Higher up the value chain, service providers that expect to use WiMax (and that will be responsible for pulling though business for the silicon suppliers) have also committed money to startup chipset developers. One such example is SK Telecom (Nasdaq: SKM) in Korea, which is working on mobile WiMax with Wavesat Inc. Other service provider investments may follow.

Startups in the space also point to Asia/Pacific original design manufacturers (ODMs), and Asia/Pacific investors generally, as more bullish on WiMax silicon than the traditional North American venture community. On the ODM side, the logic is apparently that these are the companies that will ultimately build the WiMax modems and PC cards for the brand-name suppliers, so it makes sense for them to get in on the action at the ground floor.

— Gabriel Brown, Chief Analyst, Unstrung Insider

The report, WiMax Chipset 2006 Market Outlook, is available as part of an annual subscription (12 monthly issues) to Unstrung Insider, priced at $1,350. Individual reports are available for $900. To subscribe, please visit: www.unstrung.com/insider.

mpmckay 12/5/2012 | 4:07:09 AM
re: WiMax: Hyped, but not Backed? Part of the reason for the lower level of investment, Gabriel, is as you pointed out, the Intel prsence in the CPE market. Another reason is the similarity between Wimax and existing protocols and hardware which make the new hardware more evolutionary and less expensive than the 802.11. For the same reason we'll see faster chipset deployments.
knowwi 12/5/2012 | 4:05:47 AM
re: WiMax: Hyped, but not Backed? Another possible explanation is the mountain of corporate roadkill in Bluetooth and WiFi. Every marketing weenie and CEO suggested that they would capture "only 20%" of the "huge" eventual market. VC's bought the hype and lost beaucoups bucks.

WiMAX (IMO) has been waaaaaaaaay overblown in the media, and I believe Intel's involvement to be part of the reason. WiMAX is (and will continue to be for the foreseeable future) too power-hungry and costly for reasonable deployment in CPE a la WiFi or Bluetooth (not to mention its utter uselessness in the unlicensed bands and the complications involved in using licensed bands). WiMAX is an infrastructure play, period, and the investment community is treating it as such.

It's likely a simple case of "twice bitten, twice shy".
IPobserver 12/5/2012 | 4:05:44 AM
re: WiMax: Hyped, but not Backed? Fair points. Thanks.

Your point about WiMAX as an infrastructure play is one reason to draw attention to the terminal side... to be glib, you need something to connect to.

But also there's a big debate in WiMax around how much smarts you put in terminals (smart antenna features) versus the infrastructure. Potentially, this has a big impact on cell site density, cost, end-user performance, etc, etc. WeGÇÖre sort of seeing the same thing in 3G now, where support for equalization and 16-QAM in the user-terminal chipsets is a gating factor on 3.6 Mbit/s HSDPA.

Intersting to look at the role of the terminal in all wireless technologies -- if you take 3G, WiFi, and Bluetooth, you have quite different stories in each case. To be simplistic:

WiFi -- lots of terminals, lots of usage
Bluetooth -- lots of terminals, not many users
3G -- not many terminals, not many users (yet)
Sign In