WHO WE ARE OUR TEAM WHY US CODE OF ETHICS

WIMAX OPERATOR TRACKING SERVICE AND REPORTS MARKET REPORTS CLEARSPECTRUM™ ONLINE CANADIAN INTERNET DATABASE VoIP USER AGREEMENTS

CUSTOMER PROFILE LIST OF CUSTOMERS TESTIMONIALS

ARCHIVES PRESS COVERAGE NEWSLETTERS WEBINARS

ARTICLE

Qualcomm 2007 Global Industry Analyst Event, San Diego CA, July 2007

By Robert Syputa, Senior Analyst
Contact the author at robert@maravedis-bwa.com

Qualcomm welcomed analysts to their San Diego facility to both present their position and exchange ideas.  Qualcomm and the entire wireless industry is at a point between voice domination and a data-dominated model for wireless systems and services that leads to more open business models.  We see conflicts between Qualcomm's broad technology visions and other views. Qualcomm wants to elevate wireless to more of a development platform than a particular standards issue.  This comes at a pivotal time in development of the new “personal broadband industry.”

The event started with a presentation by Bill Davidson, Sr. VP, Global Mktg. & IR, about Q's global business model.  This disclosed few surprises: Q is holding to their basic business model of supplying chips, IR licensing, and multimedia capabilities. 
A few points do stand out:

• Q stressed their embrace of 802.11n and says that their initial chips (stemming from their acquisition of Airgo) have been designed into a few embedded laptop applications and the next generation version is expected to have lower power and cost needed for PDA and phone applications. Q expects to reduce the cost below $5 per IC, which would be competitive with current 802.11b/g chips. This is in line with 802.11 industry developments.  Perhaps Q's acquisition of Airgo has resulted in a few additional design wins and, of course, Qualcomm gains new MIMO patents and design capabilities that they said have already been used to enhance their UMB (Ultra Mobile Broadband) offering, which is based on the Flarion FLASH system.
• Q brought up issues of IPR during several sessions including a presentation and Q&A session with their EVP & General Counsel, Lou Lupin.  (Recently, Lupin left following adverse rulings against Qualcomm.)
• IPR is the central issue for acceptance of UMB. Even though UMB has been modified from the system they acquired from Flarion and provides the feature set that the NGMN group has said they want, IPR focuses industry attention on Q's lock-in of technology and relatively closed nature of development.  Besides the issue of IPR royalties themselves, which can be argued as just a reasonable component of every developer’s costs, IPR focuses attention on UMB being primarily a captive development.  Although Q can claim few key contributions to 802.20 and the development of UMB systems, Q is clearly the only company willing to develop the complete chip component for a development system environment needed to deploy commercial grade systems.  More on this later.
• Lupin said that they have renegotiated most of their existing 3G CDMA IPR licenses to be "unlimited forward capture" blanket portfolio patent licenses.  I had expected this: Q had set up early licenses, such as that currently being disputed with Nokia, to be forward capture, continuing licenses.  That makes it easier for Q to migrate those licenses from CDMA to OFDM, as they are not based on individual patent considerations but on the entire, evolving portfolio and portfolios of cross-licenses.  Q wants this so that licensees get used to paying for a broad portfolio license that gets locked in, regardless of which wireless system or device they build or use. This strategy also marries well with NGMN policies: they can agree to fair licensing under NGMN and that becomes what firms have agreed to as extensions to their current licenses, which would mean that WiMAX licensees would pay as much as they would for use of UMB, LTE, or any other system.  If that becomes an accepted trend, then it eliminates a key difference between WiMAX and 3G and other OFDM developments.  Nokia is fighting this strategy, and few companies involved in WiMAX have explicitly agreed to licenses for OFDM.  However, Q has recently argued that Nokia tacitly agreed to their forward capture extension strategy by producing CDMA products. Qualcomm may try the same arguments with OFDM: by forward capture of OFDM patents into existing blanket IPR licenses, they can later argue that this amounts to tacit agreement to license their OFDM IPR under the same terms even though their CDMA technology is no longer being used.  Some cellular equipment companies may be blindly falling into Q's forward capture trap. 
• Lupin said that the recent Supreme Court ruling in the KSR case is favorable to the company. This ruling limits patents to those that meet requirements for non-obviousness to those versed in the art. Many patents have been granted for combinations of technology that may be dependent on immediate application rather than a true invention.  Sometimes, as is the case with many disputed “business methods” cases, a unique combination of technology becomes a technical or commercial advantage only when other developments occur, that is, as they become timely.  Some business methods, such as that described by Amazon's disputed “single click checkout” patent, are mere combinations of simple methods that only became an advantage when the Internet proliferated.  Many people, including me, saw the granting of that patent as ridiculous: the only reason it was not done before was because the Internet and on-line payment systems did not exist to make the obvious combination methods feasible.  The so-called “invention” was merely waiting for the opportunity to combine well-understood principles that became practical as people ventured onto the Internet.
• Qualcomm said that they will appeal Broadcom's lawsuit and the ITC trade sanction against Qualcomm will be appealed based on the Supreme Court ruling on KSR. 
• The KSR case is important to WiMAX. It will tend to reduce the importance of having large numbers of duplicative or incremental patents and will favor the essential and fundamental standing of patents.  This tends to reduce the arguments for patents that offer little innovation or improvement. 
• The MediaFlo system was rolled out with Verizon earlier this year and will see a second generation (software upgrade) rollout with AT&T & Verizon late this year.
• The company made presentations followed by a walk-through of their video production and broadcast center.  This was an impressive display of capabilities, which Qualcomm can likely extend to additional digital media efforts including sales of the broadcasting center capabilities themselves.  In fact, Q said that several TV broadcasters and cablecos have expressed interest. Key to this facility are Qualcomm-developed media servers, formatting and compiling, and command center.  The current capability is to take in direct satellite and fiber optic feeds or discrete tapes/CDs to broadcast up to 10 active channels.  At this point, Q is broadcasting 8 channels, with one additional channel being used for live-event broadcasting.
• The display quality looks very good for small format devices, QVGA-quality (320x240 pixels).  But a few analysts commented that when shown on a larger than 3.5" display, the video becomes grainy and artifacts more apparent. At issue is Q's limited bandwidth of spectrum in the N. American market, with only 6 MHz compared to Aloha's 12 MHz and up to 30 MHz that will become available in the upcoming auction. 
• I asked whether Qualcomm would bid in the upcoming auction. They said that this was being considered (in other words, they would not say).  In the closing session on Friday, COO Dr. Sanjay Jha said that Qualcomm was not going to bid on the spectrum, but this brought a comment from Q's IR that “discussions are taking place” within the company.
• At several instances, the issues of Apple's iPhone and Google's press for open access to spectrum were discussed.  These are part of the broader issue of the shift in business models between the more constrained, incumbent dominated, “walled garden” business model of the cellular industry and more open development platforms based on the emerging IP wireless environment.  Qualcomm is generally in favor of maintaining the current business model, but concedes that the shift is already occurring within the construct of the cellular industry.
• Because more applications and content must be made available to users, the walled garden approach is being pressed to keep up with customer demands.  New applications and content distribution methods, such as YouTube, can spring forward within months rather than years.  And this can cause products that do not have the flexibility to meet popular demand to lose market share. The open access model ties directly into the flexibility and ease of use of the user interface (UI).  The iPhone-type UI, which is based on Apple's OS/X v10.5, can be used to quickly bring new applications from the PC or Internet environment to the cell phone, PDA, or ultra-portable device format.
• The iPhone shows that innovation can be introduced from outside of the cell phone industry.  This refutes the argument that the cell phone industry is so dominant that it won’t face competition from the PC and consumer electronics industry for mobile broadband.  The iPhone shows that competition is expanding.
• Qualcomm stated that the royalty rates for their IPR used in handsets is less than 5%.  They did not elaborate about how this figure is determined.  Third-party studies and compilations of figures that can be determined from court cases and other sources indicate a higher figure.  The way Q stated this was for IPR used in handsets, but that ignores IPR charged for use in ICs or equipment.  IPR used in equipment was not disclosed but is understood to be higher.  And royalties for Q's IPR used in ICs adds to the total cost of IPR used in handsets.  At current issue in Nokia's lawsuit in Europe is that Nokia buys-in cell phone ICs from TI, which Nokia says exhausts Qualcomm's IPR rights.  This is called “double dipping” of royalties.
• Qualcomm says that they are in agreement with efforts underway within the NGMN group to create open disclosure of IPR and groundwork for FRAND royalty arrangements.  However, NGMN is not trying to arbitrarily set caps or value on licensing royalties.  The objective is to provide some assurance to operators and suppliers that royalties will be reasonable without being set or capped.  A major problem with standards and industry group policies is that these can be ignored, as they have only limited legal authority and they do not set limits or procedures for how IPR royalties are to be negotiated.  These policies carry weight with the courts, trade administrators, and legislators, but they often do not prevent lawsuits or trade sanctions. Qualcomm says they intend to enforce their patents and should be paid similarly for WiMAX, UMB, and LTE as is currently paid for CDMA 3G.  Nokia and others differ with Qualcomm over what are reasonable royalties for 3G, and there is likely to be a difference of opinion over WiMAX licensing as well.
• The company says that their BREW mobile development platform has gained momentum, with over $1 billion in content-publisher and applications-developer investment so far.  The overall posture of Qualcomm has naturally shifted to become centered more on development.  This shift also is expressed in the introduction of the iPhone. Exciting products like the iPhone have as much to do with the platform that underlies them as the phone itself: the iPhone takes advantage of Apple's OS X v10.5 (Leopard) windowing environment to develop a more intuitive user interface.  A similar interface can be developed based on BREW/Java or on Microsoft's WIN CE OS, but each of those will take time. New interfaces might be more rapidly developed on MS WIN CE than on Q's BREW because much of the underlying windowing software is in place.  This may be a critical difference between WiMAX and 3G wireless platforms. WiMAX, being an open IP-based system, can take advantage of the huge IT/networking and Internet software development environment, which compares favorably against the cellular industry's development environment.  Both camps will attempt to sway developers from the other. 
• Qualcomm made considerable effort to sell their EVDO rev B system development.  Rev B is an EVDO multichannel configuration.  In essence, this is increasing the number of channels, where an operator deploys base stations / channel interface cards to operate using additional bands of spectrum in high-density areas.  Qualcomm used 5 MHz channels in their demonstration.  It would take two 5 MHz channels in addition to the primary channel to provide the demonstrated bandwidth improvement of rev B.  Sprint has said that carriers simply will not go for this, as EVDO is not a symmetrical system or as efficient as WiMAX (or LTE or UWB).
• The demonstration of Qualcomm’s UMB system showed mush better efficiency and resistance to fading than EVDO rev B.  That makes Q's persistent claim that OFDM was worthwhile seem self-serving in retrospect.
• Qualcomm's CTO and Sr. VP Technology, Avneesh Agrawal, gave a presentation describing Qualcomm's technology position and challenges.  Avneesh showed slides describing the future including development of distributed wireless networks that will eventually become smart and to a large extent self-configured or "self-installed."  This is what we have foreseen for years: the development of the “smart distributed wireless broadband network” architecture.  WiMAX systems will likely usher in these capabilities before LTE appears and probably before Qualcomm can respond. 

Qualcomm has captured the number one spot from TI in sales of CDMA ICs.  This may be secure for the near future despite facing an ITC sanction against import of advanced chips containing Broadcom IPR.  Qualcomm’s overall competitive position remains strong. 

For more information you can contact the author. robert@maravedis-bwa.com