The Emergence of the Large Cap Paradigm [Part 4]
Posted November 19, 2020
This four-part series has come to an end, with George joining Richard and Steve for the finale.
Now for the last instalment. Keep scrolling…
There is no more powerful example of our emerging large cap paradigm than the continued and growing dominance of Qualcomm in wireless technology.
It’s been more than 20 years since the complete triumph of Qualcomm’s spread spectrum, bandwidth-sharing CDMA (code division multiple access) as the way to send information wirelessly, whether voice or data. By using all the available spectrum all the time, CDMA trounced rival, static “time division” systems that everyone had inherited from wireline systems. TDMA’s fatal flaw in wireless was to leave most of the wireless time slots empty most of the time, like empty railroad cars in the air.
Beyond CDMA, it’s been 15 years since — via purchase of co-founder Andy Viterbi’s Flarion corporation — Qualcomm began to disrupt its own CDMA approach with an even more demanding OFDMA (orthogonal frequency division multiple access) system. OFDMA breaks the spectrum into Fourier Transforms that find and distill multiple pure frequency carriers amid any complex waveform. Though not a Qualcomm invention like CDMA, Qualcomm came to own and master it, and OFDMA would triumph even more completely than CDMA.
And yet, the company is further ahead of all rivals than it has ever been, its reach broader, its claims to uniquely powerful intellectual property deeper.
Plus, we think the stock — up more than 6,000% from its January 1998 price ($10,000 invested in 1998 would be more than $600,000 today) — may be as good an investment today as it was then.
My big-tech champion analysts Richard Vigilante and Steve Waite think it may be even better. Who knows? They have been right many times before. I worry that Qualcomm has too many lawyers. They say that’s crucial to success today in government-regulated telecom markets. Without lawyers, such as former President Steve Altman — who framed the company’s licensing and patenting strategies — those inventive Qualcomm engineers like current CEO Steve Mollenkopf might as well return to the universities whence they came. But no doubt, lawyers and all, Qualcomm is in the broadband seat today.
Want to talk dominance?
As even casual observers of the tech world will know, there is no love lost between Qualcomm and Apple, often times Qualcomm’s best customer. The two spent years in court accusing each other of unnatural acts with patents and processes. The venom overflowed, with some $4.5 billion explicitly at stake in the lawsuit, and long-term implications several times as great.
The Apple Knife Fight
With the end of litigation in sight, Apple, sick of the internecine conflict, urgently wanted to put Qualcomm in its rearview mirror. It announced plans to buy future 5G modem chips from Intel rather than Qualcomm. (It had already been buying current generation modem chips from Intel, though Intel’s devices were demonstrably slower.)
Then Intel — the seminal semiconductor firm of all time — found it could not do the job. For its 5G modems, Apple scurried back to Qualcomm, tail firmly between legs. Apple did buy Intel’s intellectual property to build its own modem business, but is not expected to create its own modem chips for several years yet.
Maybe it will, maybe it won’t.
Now Wi-Fi Too (We Mean 6)
Q’s dominance is spreading as well as deepening. Wi-Fi modems were once rather low-tech affairs, typically single channel devices operating at low bandwidths over very short distances. It looked like too humble a business to interest the mighty Q.
Today Q dominates the field for WiFi-6 modems, operating at gigabit speeds, using breakthrough Massive Multiple Input Multiple Output (MMIMO) antennae connecting multiple devices across shared frequencies at ranges once undreamt of. WiFi6 is so robust and capacious, and above all un-licencious (that is free of government licenses and auctions), that I believe WiFi6 is going to usurp many of the markets currently contemplated for the government standard 5G.
Why? Why does this hardly young company proceed from conquest to conquest despite dozens of competitors vying for its business?
Qualcomm has dominated the field for more than 20 years because it has always addressed the twin challenges of the physical transmission of a signal, and the coding of the signal to transmit maximum information for a given amount of bandwidth as a unified problem.
Seeing the two challenges as one does not make the problem easier. Just the reverse. But it has made Qualcomm’s triumphs more complete every step of the way.
Bring in da’ Noise
Qualcomm’s answer to the problem of apparently limited bandwidth was to send signals encoded so as to be almost indistinguishable from random noise. These “white” noise-like signals, with roughly equal representation of each symbol, offered the best way to approach what is called the Shannon limit, the theoretic maximum for how much information a channel can bear.
Encoding signals as low powered noise spread across a wide swath of spectrum and decoding them at the other end, in real time, created a stunning computational challenge. But sending and receiving those faint, encoded, necessarily redundant signals also created physical challenges far more daunting than compassed by any rival system. These physical challenges were the reason CDMA was so often charged with “violating the laws of physics.”
In the late 1990s, Qualcomm claimed its CDMA technology would provide a 3X to 6X increase in transmission capacity for a given cell. This was back in G2 days, the first digital generation of digital, when users regarded 50 kilobits per second as a good connection. Yet even then most Q rivals regarded CDMA’s claims as fraudulent, so difficult to achieve did they seem. Q itself could not implement CDMA until Moore’s Law caught up and provided the computational power for the advanced digital signal processing required.
Today, single phones offering gigabits per second and cells, even at smaller scale, carrying millions of times the traffic of 20 years ago render still more fiendishly challenging both the process of encoding the signal and spreading it across multiple frequencies as well as the physics of sending or receiving it. Massive antenna arrays and signal processors orders of magnitude more powerful must compensate for inevitable echoes and route variations, frequency sharing and switching, and the ever-mounting challenge of distinguishing the signal from the noise at exponentially increasing transmission rates.
As always when demand for performance exceeds supply, Clay Christensen’s integration paradigm comes to the fore. My former partner, Clay taught me that when performance fell short of demand, and every bit of improvement counted, all the contributing subsystems needed to be modified in tandem. What mattered was not just peak performance of any module but how optimizing the performance of that module affected all the others.
I helped put Clay on the cover of Forbes with Intel CEO Andrew Grove, making Intel a standard bearer for Clay’s strategic insights on “technology disruption.”
But ever since it disrupted its own CDMA with OFDMA, no firm has exceeded Qualcomm either in disruption, or in Clay’s paradigm of integration and modularity. In the case of OFDMA, Qualcomm integrates signal processing with physical management of the channel bearing the signals. No other firm’s patents or processes touch so many critical points in the process. That is why it continues to triumph. That is Qualcomm’s fountain of youth.
Materialists, imagining that competition is the essence of capitalism, dream of “level playing fields” in which all players come to the game as equals. In truth, as our champion investor and strategist Peter Thiel is teaching the world, the goal of every entrepreneur is to create a monopoly of decisive advantages against all comers, and to extend that monopoly as far as possible.
No firm in our time has been as successful in this as Qualcomm, which is why we expect it to enjoy a more commanding position than ever over the next few years.
Throughout this series we have been proclaiming the current dominance of “sustaining innovation” to meet the limits of demand for performance at a given node in an instant of time. Only an already dominant firm can lead this sort of innovation which sustains and extends those monopolies created by the greatest entrepreneurs.
What then of disruption? Will TSMC, Qualcomm, and a handful of other current giants rule forever? If not, from where will the challenge come? From Quantum Computing? From carbon micromachines far more subtle in their application than plain old silicon. From some as yet unsuspected source?
We will deal with these questions in future prophecies.
Editor, Gilder's Daily Prophecy
Lead Analyst, The George Gilder Report
Lead Analyst, The George Gilder Report