Breakthrough Ideas (continued, written last summer, with some, but not complete, updating)
Strategies for Third-Generation Domination
Phase 2: A. The Evolutionary Strategy of Controlled Migration for 3G CDMA.
A significant factor in the success of any new product or system is whether it entails high switching costs for its existing installed base. CdmaOne networks can simply migrate to 1x networks through a controlled migration process with minimal switching costs. On the one hand, an operator’s decision to upgrade new and replacement phones to 1X can begin a process of controlled migration that, upon completion of the migration, will double voice capacity as it provides medium data rates and new services at the cost of around $3.00 per pop.
On the other hand, swapping out wireless systems entails high switching costs. It is not surprising that GSM incumbents must hold on to their installed base as long as possible to amortize the cost of the existing network and to help pay for the new one. However, the Europeans also spent $100 billion euros on two new bands, both forward and reverse, of 5 MHz spectrum for wideband 3G, meaning they must generate a ROIC that only a third-generation network can possibly bring. Thus, the Europeans are impaled on the horns of a classic economic dilemma.
Once the ITU decided on five versions of CDMA as its IMT-2000 standards, and the Qualcomm-Ericsson agreement ratified three of them, it meant the Europeans must replace the GSM Radio Air Network (RAN) because it is not compatible with spread spectrum. Europe succeeded only in holding on to GSM-MAP for its Core Network (CN), the terrestrial path from the base stations to controllers to mobile switching offices or the PSTN. The core system handles backhaul, switching, and back office functions. For GSM/TDMA networks to become 3G, it is necessary to pay now or pay later. And, the debt meter on their new and unused European 3G spectrum keeps on ticking.
Therefore, a pivotal event¾mandating spread spectrum as the standard for 3G wireless networks¾turned the tables in Qualcomm’s favor. Their first-mover 3G advantage is growing, to date producing 36 commercial 3G CDMA carrier networks in 18 countries, with many more to follow. The transition from cdmaOne networks to CDMA2000 1x is seamless and transparent to users. The infrastructure CSMs in the BTS can be swapped-out in fifteen minutes and a total 1x system can become operative within an hour or so. Older handsets remain compatible and 1x handsets can revert to the 2G design rules and parameters where necessary if the carrier uses an overlay strategy to rollout.
Phase 2: B. Separate Voice and Data Channels To Increase Spectral Efficiency.
The simplicity of having separate and optimized data or voice channels permits distinct technology development and either a unified 3G-solution of CDMA2000 1x for voice and 1xEV-DO for data or a distinct strategic path for wireless Internet and VoIP. Once data is viewed as a separate pathway to the Internet market, it introduces the possibility of new business models for both mobile wireless Internet and fixed wireless Internet.
Phase 3: Create a Universally Interoperable Global Wireless Network
Phase 3 Strategy: After using its innovative abilities to develop a universal solution that can generate worldwide network effects, Qualcomm’s Phase 3 strategic goal is to grow the worldwide market and its share of profits.
By creating the potential for a universally interoperable global wireless network, QCT’s discontinuous ZIF-simplified-RadioOne-enabled-multi-mode-architecture exponentially expanded Qualcomm’s value proposition. Specifically, QCT dissolved the 3G-Standards conundrum through an advance in performance designed to make formerly competing and incompatible systems interoperable. Qualcomm offered this advance in mobile telecommunication system integration within a new tactical set of alternative migration paths that, taken together, promise to transform what were once incompatible networks into a harmonized-world-system.
A Shift in Tactics Followed an Emergent Solution. On Qualcomm’s side, the relationship between it and the Europeans has always been pragmatic, best described as co-opetition, a dynamic mixture of cooperation and competition. Differences in Qualcomm’s presentations at this year’s and last year’s GSM World Congress in Cannes defined a sharp shift toward greater cooperation. In fact, some would say Qualcomm, wearing a white cowboy hat, is riding to the rescue of UMTS. [This was written last summer.]
In 2001, Jacobs emphasized the present availability of CDMA2000 and the problems in the UMTS timeline and GPRS/WCDMA architectures; in February 2002, Jacobs emphasized Qualcomm’s solutions to these UMTS problems. Qualcomm profits through royalties from selection of either CDMA2000 solutions or UMTS solutions, but profits more when it also sells its chips. This shift in tactics follows logically from its leapfrogging ZIF-architectural solution that facilitates universal interoperability within its chipsets.
As pragmatic as ever, Qualcomm realizes that the unity of the Europeans may never be broached because they believe their strength lies in the EU’s unity and in their local partnerships with their own national governments. However, who is to say that some company desperate for economic survival and some country wanting is own to prosper, when face-to-face with this Prisoner’s Dilemma, will not choose as its survival tactic to defect rather than cooperate.
To Europe’s chagrin, a defection from the coalition would be tantamount to shouting, “the UMTS Emperor has no clothes.” When the Emperor’s clothes cost $100 billion Euros, this difficult admission requires some Qualcomm help to save European face. Of course, given sufficient hubris, Europeans may still believe, when UMTS finally arrives, that it will be better than Qualcomm’s solutions because they are indeed the true wireless “leaders.” To me, this is more tragic than their pragmatic use of “expectations management” to stave off Qualcomm’s dominance. But, excellence demonstrated remains excellence proved, and many Qualcomm bulls believe the Emperor is naked.
Multiple Tactics Create Multiple Potential Dynamic Paths to 3G
Qualcomm’s harmonizing strategy rests on its third breakthrough idea that permitted software-as-codified-knowledge to be embodied in its integrative 6xxx chipset solutions. These new solutions combine modes, bands, and network into a seamless, integrated, interoperable mobile network of networks. Each combination gives you the value of two networks for the price of one and keeps on adding new modes until you have a global mobile network that is joined to both the networks of PSTNs and the Internet. Not only two for one in reach, but also it promises an increasing richness of potential applications as it taps the world-market of developers. Increasing functionality drives consumer demand. Always on, always with you global connections expand the size and value of the telecommunications pie. If the principle of increasing returns, implies this corollary, of networks, there shall be few, then one universal network sounds like the Goldilocks size.
Upgrading to 1x is a no-brainer for an existing IS-95 network because it is easy, fast, cheap, and pays for itself by doubling voice capacity as it simultaneously introduces customers to the value of selected data application, like streaming media, games, picture-swapping, and PIM. Also, it is transparent to end-users because it only requires that an operator stop offering IS-95 handsets and instead offer 1x handsets as the generic CDMA handset. In the second half of 2003, ZIF architecture will enable better price/performance in entry-level or low-end handsets, the MSM6000, with 1x rel 0 and14.4 kbps data and 2-way SMS, or the MSM6050, with 1x rel A, with optimized gpsOne, 153 kbps FL/RL, basic multimedia features, than the IS-95 models anyway. This natural migration means that you do not have to sell either the ideas of 3G or new data applications in advance of purchase because demonstrating excellence proves excellence to customers, who naturally enjoy the new features of mid-tier handset with the MSM6100 with colors, position location mapping, multimedia, and cameras, and tell their friends, “try it, you’ll like it.”
Qualcomm’s harder job is to ease the way for previously incompatible networks to upgrade as painlessly as possible. Each operator that makes the switch is a testimonial to the shift in mindshare from GSM/UMTS leadership to Qualcomm as the third-generation architectural leader. If that battle for mindshare were already won, it would be excellent for Qualcomm but less good for prospective investors.
Playing the Asia card and demonstrating excellence to prove excellence remain effective tactics, but also Qualcomm has several new tactics oriented toward potential GSM/UMTS markets: (a) to sell its own WCDMA/UMTS chips competitively, (b) to solve UMTS’s asynchronous search problem through partial or complete synchronization, (c) to overlay GSM with GSM1x, (d) to sell 1xEV-DO independently or as a follow-on to GSM1x, (e) to compete in the 450 MHz market in underserved Northern or Eastern Europe, Russia, and former Soviet Union countries; and (g) to compete for migrations from stranded TDMA.
Each of these tactics is secondary to this Phase 3 architectural strategy: Rapidly deploy multi-mode MSMs in 2003 to compete as the standardizing crucial interface¾a world phone platform, that, taken together, will produce a single, seamless, global 3G-spread-spectrum mobile wireless market.
Establishing Context: The Times, They Are a Changin’. The media tell the story daily of GPRS problems and UMTS delays in Asia and Europe. Because of the “religious” war, expectation-managers spin these stories. For the investor, the only remedy for reducing residual uncertainty is to make the best possible analysis and form your own opinion. In fact, this may be the informed individual investor’s greatest advantage because Wall Street, as yet, understands neither the breakthrough technological advantages of QCT’s architecture, nor Qualcomm’s business strategy, which is based on its unprecedented degree of architectural control over a huge market for the next ten years or more. To invest in Qualcomm, you should believe that Dr. Jacob’s is trustworthy. Until that trust develops, just watch him, he will prove it.
The facts will become increasingly clear within the next couple of years, clarified by the GPRS failures and UMTS delays that contrast with the CDMA2000 1x and 1xEV-DO rollouts or the RadioOne successes. Time will reveal that an enormous shift in competitive advantage was created when, to advance 3G requirements, spread spectrum architecture became the only architectural possibility for supplying a wireless infrastructure capable of meeting the growing demand for spectral efficiency, in general, and data, specifically. This required Qualcomm’s spread spectrum intellectual property. Its continuing advance still demands its innovative abilities. Its spread spectrum and radioOne innovations sustain its competitive advantage and deepen its strategic control.
What does Dr. Jacobs’ claim that changing times will tell? My incisive paraphrase is: Europe has oversold GPRS and EDGE on both its time-availability and supportable data rates. Both approaches were intended to provide always-on packet data connections, but the underlying GSM architecture inherently limits their success to a low speed, relatively expensive approach to providing data. This is so because data channels overburden already limited voice capacity. This is so because data-over-GSM becomes even more susceptible to interference than its GSM parent designed for voice only. This is so because GPRS demands excessive power for handoffs. This higher power places their already limited frequency reuse pattern in further jeopardy, requiring either a much lower data rate or exponentially more base stations with ever more complex frequency reuse plans. Absent the possibility of an elegant CDMA solution, GPRS simply increases transmitting power (and power consumption). Not only does this failed solution drop data calls, reduce coverage, consume voice capacity, and produce hot phones that deplete batteries and emit radiation, but also it adds infrastructure, increases operating costs, cuts margins, pressures income, and limits ROIC.
Nevertheless, as voice matures, mobile data must increasingly drive the growth of revenues and earnings. These inherent problems in GSM architecture as a medium for delivering data were the very reason that the European’s recognized and accepted the need for a worldwide switch to spread spectrum architecture for 3G.
If he were being candid rather than diplomatic, Jacobs might add, GPRS appears poised on the precipice of technology failure, and EDGE will never be. If being boldly candid, he might add, Qualcomm, and only Qualcomm’s platform, can continually improve any version of 3G-spread spectrum architecture, including the flawed and unfinished UMTS. Deny as you might, the times are changing. And, indeed, time will soon tell.
Compete in UMTS. Currently, Qualcomm is the world leader in ASIC product development for WCDMA/UMTS. In September 2001, it delivered on-time worldwide sampling of a UMTS solution, called the MSM5200. Last December, the first international call was made using its UMTS-compliant test phone. Qualcomm’s UMTS Test Mobile became the de facto standard. It now includes both UMTS and GSM/GPRS in a tightly integrated chipset. Previews and demonstrations for the majority of European carriers were completed and followed by complete IOT testing with the major carriers and eight infrastructure vendors.. At the GSM World Congress in February 2001, live network calls were placed by Ericsson, Lucent, and Nortel using Qualcomm’s MSM in its buttoned-up, small form factor test phone. Also at the GSM World Congress, one vendor demonstrated UMTS streaming video (64 kbps circuit switched data) using the MSM 5200. In March, a wireless GSM call was demonstrated at CTIA. Qualcomm experienced no difficulty in meeting all GSM interoperability standards. In April 2002, Qualcomm was the first company to sample direct conversion architecture, and Qualcomm also achieved the carrier-required 384 Kbs for UMTS mobile packet data calls with six infrastructure vendors..
First to market with UMTS, five OEMs are already committed to using Qualcomm’s chipset in new handsets, with more on the way. In addition to the MSM5200, the MSM 6200 began sampling on June 17, three months early. Just in time for J-Phone of Japan (a joint venture of Japan Telecom and Vodaphone), who announced it UMTS plan to use Sanyo handsets, which use MSMs from Qualcomm. J-Phone scheduled trials for July, but they were pushed back. On May 31, although customers are still waiting, to differentiate itself from NTT’s FOMA, J-Phone began touting the worldwide compatibility of its Sanyo phone (I believe using the MSM6200) that will enable multi-mode UMTS/GSM/GPRS roaming in 179 countries internationally.
At the May Analysts Meeting and again in June, Don Schrock reported strong international interest in the multi-mode single-chip solutions for both UMTS/GSM and GSM1x. The MSM 6200 is a multi-mode UMTS/GSM/GPRS RadioOne MSM solution, which includes proprietary ZIF-baseband processors to receive, transmit, and manage power. By reducing the BOM and main-board-size by 50%, it raises architectural and cost challenges that Nokia’s planned UMTS chipsets, scheduled for September (although not released as scheduled), cannot equal.
In 2003, the MSM6250 and the MSM6500 will begin sampling. The MSM6250 segments the UMTS market into mult-mid and high tier segments because it adds a newly developed Launchpad feature set: MPEG4 video encoding and decoding; megapixel resolution JPEG still picture camera; dedicated graphics accelerator,for advanced gaming, built-in camera sensor interface; wireless assisted and stand-alone GPS position location, ARM 926JE process with Java hardware accelerator, high quality 40-polyphony MIDI sound synthesizer, MP3 player and intergrated stereo Hi Fi audio DAC, and an integrated WAP 2.0 browser to support MMS-based applications. The MSM 6500
segments the GSM1x market into a multi-mid and multi-high tier market by combining GSM/GPRS with both CDMA2000 1x and 1x EV-DO networks. The Europeans have no current or anticipated competitor for high data rate and cannot match Launchpad’s advanced features.
If the Europeans stonewall purchasing these chipset solutions, will the rest of the world docilely follow their leadership? Not likely when evidence from competing networks proves Qualcomm’s case. At the May Analysts Meeting and again in June, and continuing into 2003, Don Schrock reported strong international interest in the multi-mode single-chip solutions for both UMTS/GSM and GSM1x. Qualcomm appears to believe that European adoptions are possible. |
