Monday, December 28, 2009

Len, we hardly knew ye

On the day before Christmas, Qualcomm announced that its COO Len Lauer is leaving to become CEO “at another company.” Although his hiring was a big deal three years ago, I didn’t see him quoted a lot during his time as group president and later COO.

I kept waiting for someone to leak what that company is, but no one has said — and he refused to say when the UT asked him Thursday. (My hunch is that the former CEO of Sprint Nextel is not leaving to head a small startup.)

A few things bother me about this announcement. I agree with Sebastian Rupley of GigaOM:
I know companies don’t typically release news a day before Santa’s visit, unless they want to push things under the carpet.
He’s right, but no one is going to brag about losing a top exec and so that part’s understandable.

I’m only slightly bothered that Lauer was one of the troika of leading execs at the Dec. 16 annual Qualcomm “town hall” for the local telecom community — and then 8 days later he’s gone. It’s also curious that the December 2006 press release announcing his appointment has been deleted from the US website (but not the Russian site or the Google cache).

More seriously, one of the things is that we’re seeing the replay of what happens at any big company where there’s no chance of becoming CEO — the most experienced execs are leaving. Then-COO Sanjay Jha in August 2007 left to become Motorola’s handset CEO (when/if it’s spun off).

Yes, it’s a testament to the quality of the people at Qualcomm, the experience they gain and the value that other firms place on that. (Of course Lauer had excellent operating credentials before joining Qualcomm.)

However, when your CEO is 47 and the scion of one of the founding families, it’s obvious that it’s going to be a long, long time before anyone else gets to control the reins. I haven’t seen any evidence that Qualcomm is prepared for its new role of providing executive talent to the rest of the industry.

It would be good if the board were populated with someone who lived through the two decade Jack Welch era at GE, where the company groomed talent and saw them bail out in despair of ever taking the gold ring.

Perhaps some of the GE experience can be used to keep good talent longer. In retrospect, some of the GE CEO wannabes — think Nardelli — turned out to be spectacularly bad CEOs without the GE infrastructure behind them. Qualcomm has a unique position of strength in an industry that still is growing (at least a little), and it may hard to be a huge success in a company or industry that poses more serious challenges. (Exhibit A: Ed Zander.)

There was one other troubling thing about the announcement. With Lauer’s departure, Qualcomm is juggling responsibilities between people to fit the men (they’re all men), just as they did when Lauer was appointed 33 months ago.

I’ve never been an executive in a Fortune 500 company, but it seems to me that the lines of the various divisions should be consistent and not re-juggled based on who’s available this week to run them. On the one hand, QCT head Steve Mollenkopf has the operating experience most relevant to running (as he now will) Qualcomm MEMS Technologies, makers of Mirasol.

On the other hand, will running QCOM’s biggest division cause the QMT growth possibilities to be neglected? And if this was such a good idea, why wasn’t it done earlier? Lauer’s service business were assigned to president Steve Altman, and it appears the top rungs of the executive team also includes Andrew Gilbert Derek Aberle (head of QTL, historically Qualcomm’s most profitable division.)

I’m guessing that another re-assignment of responsibilities will come in the next 12-15 months. Hopefully, that will be as the result of substantial growth from one of Qualcomm’s newer divisions (like QMT or MediaFlo) rather than more juggling to fit the changing personalities in the executive suite.

Updated Dec. 30: As a Qualcomm reader points out, Derek Aberle is head of QTL while Andrew Gilbert is head of Qualcomm Internet Services.


Update: see this report of Len’s new job.

Friday, December 18, 2009

Qualcomm: beyond the cellphone

CEO Paul Jacobs and two of his top executives were on stage Wednesday for the annual San Diego Telecom Council CommNexus “Qualcomm Town Hall”. Deutsche Bank analyst Brian Modoff pitched his own questions, as well as ones texted to him by audience members.

Bruce Bigelow covered the story for Xconomy San Diego. (At the reception beforehand, I chatted with Bruce — we’d never met although I’d traded emails with him when he was at the U-T.)

Bruce’s story focuses on one of the major themes for Qualcomm nowadays, which is the ongoing shift from voice to data. He emphasizes discussion of 4G deployment, although the 4G questions were all about LTE (particularly Verizon’s rollout starting next year) rather than WiMax. And indeed, looking at my notes, 4G/LTE was one of the major topics of the evening.

As Bruce notes, Jacobs noted that the 4G rollout will happen over time. One question relayed by Modoff: analysts are asking was when will there be single mode LTE devices? The answer was that there will be multimode devices as long as analysts have a job on Wall Street — and Jacobs added “hopefully that's not because they have a short career.” Jacobs also talked about limited prospects for improving spectral efficiency with 4G, something he and his dad had discussed in their rare joint keynote at CTIA in October (also held here in San Diego).

Still, in listening to Jacobs talk, I found at least one answer to my longstanding question: how will Paul be different than Irwin?

The short answer: Irwin is about radio and Paul is about mobile computing.

Yes, Irwin was UCSD’s most computing-oriented professor at the APIS department back in the 1960s — to the point that the department chair despaired of losing him when he left on leave in 1971 to take charge of Linkabit. However, Irwin’s publications and influence were as a popularizer and entrepreneur for Shannon theory, including his greatest career accomplishment — turning CDMA into a business.

And certainly Paul has his Berkeley EE PhD and can talk spectral efficiency with the best of them. But his career and his passion seems to be as a device geek, and Wednesday’s session seemed to emphasize that he is taking Qualcomm not just beyond voice, but also beyond the cellphone. (Of course, he was also born into the software/microcoding era, and has fully internalized software-based device design to a degree that perhaps Dad never could).
Indeed, the idea of Qualcomm beyond cellphones was a recurring theme during the executive Q&A. Beyond LTE, two other topics that came up over and over were the Snapdragon chip and the FLO TV 700 Mhz broadcast television service.

Prompted by Modoff, QCT president Steve Mollenkopf kept coming back to Snapdragon over and over again. It’s not clear how clear how successful the Qualcomm ARM cpu/communications chip will be, but it was very clear that pushing Snapdragon is one of Qualcomm’s major strategic thrusts for 2010 as it moves into mobile devices such as “smartbooks”.

Mollenkopf was asked to compare Snapdragon vs. Intel’s Atom. He made it clear that they’re in different niches now, but eventually the two products (and presumably companies) will be directly competing. Indeed, it’s hard to imagine how the Snapdragon smartbook will be distinct from an Atom netbook in 2011 or 2012 — other than the choice of operating system (Android or perhaps Windows CE vs. genuine Windows).

The other non-cellphone device that came up was the Personal TV portable device that Qualcomm is selling to promote use of FLO TV in the US. COO Len Lauer pulled it out more than once — and also mentioned that it was available at Best Buy, Amazon and Radio Shack. His ongoing plugs for the PTV became a sort of running joke for the evening.

It was not clear where the PTV is an attempt to bypass cellphones to win FLO TV adoption, or merely to create a buzz and urgency while waiting for cellphone makers and the Big Two (AT&T and Verizon) to more enthusiastically promote the service.

While they are less about radios and more about mobile devices, the PTV and Snapdragon also reinforce an existing Qualcomm strength dating back to Irwin (and Andy Viterbi’s) earlier company: Linkabit. It is clear that Qualcomm was born (and remains) a systems company, designing things from end to end rather than individual components. Yes, it may partner for other pieces of the value chain, but it can also bring products to market if that’s what it takes (CDMA cellphones, the PTV).

I think this is one area where Qualcomm is quite different than Intel: it is and has always been a systems company, while Intel has spent 30+ years seeking R&D and scale economies from high-volume semiconductor manufacturing. Given their complementary strengths, and overlapping customer base they could conceivably work together.

However, given their size and corporate egos, I think the two companies are destined to remain more competitors than coopetitors. To Qualcomm’s credit, its relations with the third titan in this segment (Nokia) are somewhat better than Intel’s.

Personal note: This was the first time I’ve attended an evening event in San Diego and then flown home afterwards. At the Southwest terminal in Lindbergh, I was relieved to see Modoff and an associate also there after the talk, catching a 9pm flight to SFO as I caught a flight to SJC.

Monday, December 14, 2009

"First" 4G networks

The news Monday reported that TeliaSonera launched “the world’s first fourth-generation wireless service” (as the FT put it) Monday in Stockholm and Oslo.

The statement is untrue, since the industry (including vendors) considers both WiMax and LTE are 4G services, and Sprint rolled out WiMax in 2008, and Clearwire is continuing the WiMax launch across the US. (I’m a bit of a WiMax skeptic, but fair is fair: they did get there first.)

Information Week more accurately calls it “the first LTE deployment”:
The first LTE deployment gives TeliaSonera first-to-market bragging rights as several other global service providers prepare to roll out LTE service. Initial deployments are expected to be limited to laptops with modem cards although handsets with voice capability are expected to be introduced later.
Verizon has been running US trials of LTE and hopes to launch the service next year.

Qualcomm claims patents on both LTE and WiMax. They are likely to upgrade existing W-CDMA clients to LTE, and they are hoping to collect royalties from WiMax as well.

Saturday, December 5, 2009

Odd blast from the past

My Google News watch on Qualcomm today turned up this interesting news story:
Qualcomm Incorporated (NASDAQ:QCOM) Appoints Rich Sulpizio as President and CEO

Ft Lauderdale, Florida 12/05/2009 12:55 AM GMT (TransWorldNews)

Qualcomm Incorporated (NASDAQ:QCOM) is welcoming Rich Sulpizio as its new President and CEO. Mr. Sulpizio has previously served the company as its President and COO. His new responsibilities will include the provision of strategic and operational leadership to grow the company’s market share worldwide.
No, Paul Jacobs isn’t suddenly unemployed: he still has his day job.

Instead, a more accurate account is given here:

Qualcomm Appoints Rich Sulpizio President And CEO Of Qualcomm Enterprise Services
12/2/2009 4:20 PM ET

(RTTNews) - Wednesday, Qualcomm Inc. (QCOM: News ) appointed Rich Sulpizio as president and chief executive officer of Qualcomm Enterprise Services. Sulpizio rejoins Qualcomm with more than 10 years of leadership experience in the Company, including previously serving as Qualcomm's president and chief operating officer. In his new role, Sulpizio will provide strategic and operational leadership to grow QES' market share globally. Sulpizio will succeed Bob Walton, who has decided to look for other opportunities.

Sulpizio most recently served as president of MediaFLO USA Inc., a wholly owned subsidiary of Qualcomm, overseeing the development and deployment of MediaFLO technology, and previously as interim president of Qualcomm China. From 1998 to 2001, he led the Company as president and chief operating officer.

Qualcomm has only had two CEOs, both named Jacobs. However, in 2000-2001, Qualcomm was discussing spinning off the QCT division, with Sulpizio nominated to be CEO of the firm dubbed “Spinco”. In 2000 — as in 2006 — part of the motivation was to avoid giving 3G IP holders a hostage to take in bare-knuckles negotiations over QTL royalties.

Another explanation I heard, however, was that the split would allow Paul Jacobs to become CEO of QTL while the older and more proven Sulpizio would take QCT. Instead, Jacobs fils became CEO of the whole enchilada back in 2005.

I have not had a chance to interview Sulpizio yet. His decision to come back appears unusual, particularly since Paul Jacobs appeared to have cleared out almost any executive associated with his father’s leadership. (CTO Roberto Padovani being a notable exception.)

Certainly I find it encouraging that Sulpizio and Jacobs have overcome whatever natural rivalry they had a year ago — perhaps encouraged by the board, which still includes some faces who worked with Sulpizio in his earlier QCOM stint.

Tuesday, November 24, 2009

Paul Jacobs negotiates way out of EU complaint

The European Commission today announced that it’s dropping the four-year-old antitrust case against Qualcomm‘s patent licensing policies. Don Rosenberg — general counsel of Qualcomm since October 2007 — was understandably pleased:
Qualcomm is extremely pleased to learn of the European Commission's announcement. After an extensive four-year investigation into Qualcomm's practices, and despite the coordinated nature of the complaints made against it, the Commission has terminated its investigation with no finding of a violation
The Case

The complaint was filed in October 2005 by two European telecom makers (Nokia, Ericsson), two Japanese makers (NEC, Panasonic) and two US competitors for Qualcomm’s chipsets (Broadcom, TI), sending Qualcomm’s shares down 4.6%.

Since that time, Qualcomm settled with two of the original complainants. Its biggest customer — Nokia — renewed its license to Qualcomm’s patents in July 2005 on terms believed to reduce its long-term royalty payments. After losing repeated lawsuits with Broadcom, last April Qualcomn settled all IP complaints between the two companies. Both parties dropped out of litigation against Qualcomm.

In anticipation of the commission’s decision, Ericsson on Tuesday dropped its complaint to the EC. (Note: this is Ericsson the leading supplier of cellular infrastructure, not Sony Ericsson, the handset also-ran). Ericsson claimed victory in its attacks on Qualcomm’s business model due to a $208m fine in Korea and an ongoing investigation in Japan.

Since 2005, TI and Qualcomm have avoided litigation, but in 2007 Qualcomm passed TI to be the top mobile handset chipmaker. Qualcomm is on track to be the 6th largest semiconductor vendor (albeit a fabless one) this year.

The official EC statement said
The Commission committed time and resources to this investigation in order to assess a complex body of evidence, but has not as yet reached formal conclusions.

All complainants have now withdrawn or indicated their intention to withdraw their complaints, and the Commission has therefore to decide where best to focus its resources and priorities. In view of this, the Commission does not consider it appropriate to invest further resources in this case.
The Commissioner

One possible factor in the decision may have been the expected end of the term of the EC’s Competition Commissioner, Neelie Kroes, who saw her role as remaking the global world order for the benefit of EU firms.

Kroes has openly campaigned for more aggressive oversight against US market leaders (Microsoft, Intel, Qualcomm). She unilaterally imposed a $1.4b fine on Intel for its actions against its US competitor. She hoped to use her success over Microsoft to bring down Qualcomm as well.

Two months ago rumors surfaced that Kroes wanted to convene a special panel to examine Qualcomm’s activities. At the time, some speculated that others within the EC opposed Kroes because there wasn’t much of a case. Tuesday, Kroes won reappointment to the European Commission, but is widely expected to be rotated out of the powerful Competition position.

The CEO

I believe that clearing up the uncertainty and risk associated with these various legal threats constitutes Paul Jacobs’ greatest accomplishment thus far as Qualcomm CEO. It’s tough to follow an industry legend — the only CEO the company had ever known — who also happens to be your father. (Note: for my book, I’ve interviewed Irwin Jacobs twice but have only heard Paul Jacobs give speeches.)

However, Paul Jacobs inherited these problems, the inevitable collision between Qualcomm’s take-no-prisoners royalty model and the decision of its GSM rivals to use CDMA technology in their 3G standard.

When Qualcomm (charitably) had mixed results in the courtroom, its new CEO took a more conciliatory approach with its most determined opponents (Broadcom and Nokia) to create a win-win outcome. (I’m betting Qualcomm also worked out less public accommodations with its other adversaries). The resolution of these cases brought the end to the EU case, the most serious threat to its business model.

When he ascended to the CEO position in July 2005, Jacobs was best known as a technology enthusiast who emphasized new revenue streams such as MediaFLO. The jury is still out on these businesses, but given his pragmatic approach towards litigation, one can hope he will take a dispassionate look at them and decide to pull the plug if they’re not panning out.

Wednesday, November 11, 2009

Motorola to spit out General Instruments?

One of the first and most important Linkabit spinouts was the Videocipher division, which designed and built settop boxes for 10' C-band dish owners who wanted to watch HBO. In the late 1970s, Videocipher was one of Linkabit’s fastest growing divisions and slated to become its first high-volume consumer business; the Linkabit founders have said that the capital needs of building this business were a major factor in selling the company to M/A-COM.

M/A-COM acquired all of Linkabit in 1979-1980, and then after Andy and Irwin left, M/A-COM sold Videocipher (and other commercial TV businesses it had acquired) to General Instruments in 1987 for $220 million. The local San Diego GI team worked on a number of issues related to settop boxes, including the DOCSIS standard that both enabled the widespread adoption of cable modems and commoditized the business, destroying profit margins.

In 1999, Motorola offered $11 billion for General Instruments and the deal closed in January 2000: for almost a decade, the local ex-Linkabit outpost has been a Motorola facility. Continuing to build on the original Videocipher direction, San Diego is headquarters for the public key infrastructure (cryptography) operations of Motorola Inc., assuring Hollywood today (as 30 years ago) that no one is going to get copyrighted entertainment without paying for it.

Now the WSJ reports (picked up by others) that Motorola is looking to unload the settop box division for about $4-4.5 billion. It’s not clear whether the buyer would be another telecom company or a private equity firm hoping to spruce it up and resell it. I’ve not been out to the Sequence Drive facility in several years, but my impression was that the division (like the rest of Motorola) has been enduring a series of job cuts.

Motorola — co-led by former Qualcomm COO Sanjay Jha — is also said to be re-opening plans to dump its declining handset business, once the market leader but for the past few years in rapid freefall. Jha was hired in August 2008 to run the handset business, but the spinoff was halted this time last year after it was clear the business had to be fixed before it could be sold.

As Jha predicted six months ago, Motorola has gotten a nice bounce from its Android phones, suggesting that now is the time to sell (or spinoff) the division. No word on whether Jha will stay with the mobile division he currently heads when it’s spun out; in the past year he’s been much more visible than the other CEO, Greg Brown, who heads the broadband division.

Thursday, October 22, 2009

Despite Qualcomm, Sprint losing Kindle

Earlier this month, Amazon introduced an “international” (i.e. GSM) version of the Kindle 2 that uses AT&T (like the new Barnes & Noble Nook). Now CNET reports they’ve cancelled the CDMA Kindle 2, leaving only the 10" model on the Sprint EV-DO network (for now).

To get it all straight, I had to do some digging. Amazon has thus far released 4 Kindles:
Sprint was clever to partner to get this business, so it’s a shame they’re losing it — particularly since the Kindles began life as Powered by Qualcomm.

The original Kindle was powered by a Qualcomm MSM chip, and according to an iSuppli teardown, the Kindle 2 included a MSM6801A as part of a Novatel Wireless card.

A purported 2008 internal Qualcomm news article credits Qualcomm with working hard to launch the original Kindle:
The device is the result of a two-year collaboration with Qualcomm Enterprise Services, Qualcomm CDMA Technologies, Corporate R&D, Corporate Product Security, the Office of the CTO and Corporate IT.

Amazon was originally introduced to QES through a modem manufacturer who hoped to provide modems for the Kindle. The same company had also worked with QES on the CardioNet project and knew that the wireless connectivity that Amazon envisioned for the Kindle was exactly what QES could deliver.

“The project was a natural fit for QES. We have been delivering end-to-end wireless data services for almost 20 years,” said Susan Hennenfent, Senior Director of Product Development for QES. “Our expertise allows our customers to focus on their businesses while we do what we do best.”
I haven’t seen a teardown on the Kindle 2 International to see whether Qualcomm kept the business or was pushed out of the BOM.

For phones, dual-mode Qualcomm chips have allowed Verizon (and occasionally Sprint) to support global roaming for voice and 3G data. I’m guessing that Amazon decided to switch to GSM rather than pay a premium for the dual-mode chip. Price has been a major concern: the latest Kindle 2 is $259, exactly matching the Nook and a full $100 below the first (CDMA) Kindle 2 when introduced 8 months ago.

Monday, October 19, 2009

End to CDMA iPhone hopes?

Rumors have been extant all year that — after turning down the iPhone back in 2006 — Verizon Wireless was actively negotiating to get the iPhone on its network. (Cynics noted that even the appearance of talking helped both Verizon and Apple against AT&T).

Verizon is locked in a battle for dominance with AT&T, and its ads this month have been making fun of AT&T’s network reliability with a pun (“there’s a map for that”) that also attacks the iPhone.

Now a new ad campaign has the Verizon promoting the Android-based Motorola Droid by attacking the iPhone. As John Murrell of Good Morning Silicon Valley (of the SJ Merc) wrote
This weekend saw the launch of a TV commercial and a Droid teaser site that opens with a scrolling list of direct jabs — “iDon’t have a real keyboard. iDon’t run simultaneous apps. iDon’t take night shots. iDon’t allow open development. iDon’t customize. iDon’t run widgets. iDon’t have interchangeable batteries.” — and finishes with a hard right: “Everything iDon’t, Droid does.”
Murrell suggests:
[T]he direct Droid attack would seem to be more evidence that Verizon has dropped any hopes of landing an iPhone deal itself and has chosen to cast its lot with Google
although he hedges his bets by noting new rumors of Apple/Verizon cooperation on a CDMA/LTE phone.

It’s pretty clear that Apple won’t develop a CDMA version without Verizon. I believe Sprint has too many problems to be a major launch customer, and it’s already put big eggs in the Palm and Android baskets.

Yes, China Telecom and KDDI (in Japan) together have as many subscribers as Verizon Wireless. However, the iPhone demand in those two countries is much weaker than in the US and Western Europe, so if there isn’t a US CDMA iPhone, then there isn’t going to be one.

Tuesday, September 29, 2009

CDMA growth in China

The FT this morning has a profile of the innovative management of China Telecom, the CDMA carrier in the PRC.

From what the article says, the Ministry of Information Industry's grand reorg of the Chinese telecommunications services sector — reallocating mobile licenses and consolidating weak operations with strong ones — seems to be having the desired effect. Competition is intensifying and firms are investing heavily in rolling out 3G infrastructure, both depressing profits.

In the reorg, China Telecom inherited the 2G CDMA operations run by Great Wall (a People’s Liberation Army affiliate) that became part of China Unicom. CT has been rolling out EV-DO, but many assumed that with its US technology it would be at a disadvantage compared to China Unicom (with GSM/W-CDMA) and China Mobile (the dominant carrier, using China’s home-grown TD-SCDMA).

However, the FT says that China Telecom has the highest growth of the carriers, jumping from 28m to 44m subscribers. (By comparison, China Mobile has 500m). To quote FT:
Analysts say China Telecom has established itself as the best run of the three telecoms operators.

“They are much more of a pragmatic, commercial animal than the political animal of China Unicom,” says Duncan Clark, chairman of BDA, a telecoms consultancy in Beijing.
The article credits the company’s research lab (China Telecom Shanghai Research Institute), and a company-wide innovation initiative that encourages workers to be more innovative in developing and packaging mobile services. The company runs an annual internal innovation contest, while engineers are sent to the provinces to work with regional marketing reps to understand the market.

Obviously it will be a while before China Telecom threatens China Unicom (142 million subscribers), and it seems unlikely to directly challenge China Mobile. Still, Qualcomm (and the global cdma2000 ecosystem) are fortunate to have drawn such an innovative ally in the world’s largest mobile market.

Thursday, August 27, 2009

FUD against EV-DO in China

W-CDMA supporters in China are running a campaign of fear, uncertainty and doubt (FUD) against EV-DO efforts there, in a rehash of their strategies from the 2G CDMA wars.

In the great 3G reorg of China last year that created three mobile operators, China Unicom got GSM/W-CDMA, China Telecom got CDMA/cdma2000 and China Mobile got the homegrown TD-SCDMA. On January 7, all three were awarded their 3G licenses after more than five years of delay.

In a story on China Unicom’s plans to sell the iPhone in China, Forbes included some commentary on China’s three-way fight for 3G market share:
Unicom was given a WCDMA (Wideband Code Division Multiple Access ) license. China Mobile, which holds a commanding 74% share of domestic subscribers, obtained TD-SCDMA (Time Division Synchronous Code Division Multiple Access), a homegrown 3G standard based on spread spectrum technology that is different from approaches used in the West). And China Telecom was allowed to develop CDMA 2000 (a hybrid 2.5 or 3G technology which built on code division multiple access, or CDMA, technology used worldwide.

Among the three, China Unicom's WCDMA is regarded as the leading 3G approach. What's more, it's the only 3G technology used by the iPhone.
This sort of FUD marketing strategy is a classic technique by a dominant player to marginalize competitors. IBM in the 1970s was credited with inventing the strategy for the tech industry, and of course Microsoft used it against open source software or other rivals in the 1990s.

Perhaps cdma2000 1x is a 2.5G technology ala EDGE. The fights between GSM and CDMA camps over what should count as “3G” caused a fair amount of disagreement.

But certainly EV-DO — as fast as W-CDMA’s HSDPA — is a 3G technology. This year China Telecom has issued a series of tenders for procuring EV-DO equipment. The company is preparing a new round of handset tenders and can brag about superior data performance.

So to say China Telecom is only developing “2.5G” is a lie. The claim about W-CDMA being the leading technology is true from an adoption sense, if not from a standpoint of technological performance.

I wasn’t there, so I don’t know where the reporter got the bad information. The Chinese equipment suppliers are the least likely suspects, since leading firms like Huawei and ZTE provide equipment to all three of the major standards: W-CDMA, cdma2000, and TD-SCDMA.

However, based on the bitterness of the GSM vs. cdmaOne (or W-CDMA vs. cdma2000) fight, the most likely suspects are those who want to commercially weaken cdma2000 and China Telecom. That leaves either rival operators (China Unicom, China Mobile) or foreign W-CDMA equipment providers such as Nokia and Ericsson.

Tuesday, August 18, 2009

InterDigital loses key patent case

On Friday, InterDigital lost an effort to enforce four key 3G patents on Nokia’s cellphones.

Of course, IDCC will appeal. Company supporters hope IDCC can win a favorable settlement — as it did with Samsung.

I don’t see a settlement in the cards. As the world’s largest cellphone maker — and thus the largest royalty target — Nokia has been playing hardball all the way. Of course, in 2008 Nokia settled its war with Qualcomm, on very favorable terms.

My sense is that Nokia places a much greater value on the Qualcomm patent portfolio — or at least considers it impossible to work around all their IP without getting some sort of license.

In the face of a precedent that (if sustained) could begin to destroy its business model, IDCC shares lost 23% of their value on Monday and were essentially flat Tuesday. This seems a little optimistic.

Barron’s reports:
Hilliard Lyons analyst Tom Carpenter maintains his bullish stance on InterDigital despite the setback at the ITC. Carpenter notes that Nokia could still lose on appeal, which could trigger a licensing deal between the two companies. He thinks Nokia could still end up paying IDCC 35 cents per handset, or as much as $300 million over five years.

Even if the company fails to win on appeal, Carpenter contends IDCC’s patent portfolio could make it a potential acquisition target for Research in Motion (RIMM), Cisco (CSCO) or even Nokia. The potential appeal: IDCC holds a large number of patents on LTE, the 4G transmission technology that has been adopted by most wireless carriers.
While I understand the large pool of IDCC bulls, I don’t get this, either. I guess it all depends whether you consider this court case a fluke, or representative of the quality of the rest of its patent portfolio.

Friday, July 31, 2009

MetroPCS latest perpetrator and victim of commoditization

On Thursday, MetroPCS cut its price for unlimited everything to $45, vs. $100-130/month for the major carriers. Its price for voice text and web is only $40.

Also on Thursday, shares of MetroPCS fell 11% while rival Leap Wireless (aka Cricket) fell 8%, in the face of increased competition from Sprint’s Boost Mobile and TracFone.

The big news this week is that Sprint is buying out the partners in its largest MVNO, Virgin Mobile. Is this good for MetroPCS and Leap?

Quoting one set of analysts, the WSJ M&A blog said yes:
UBS said:
“Transaction helps rationalize US unlimited prepaid market. This transaction helps rationalize the unlimited prepaid market in the US, which had grown to 5 players with Tracfone’s Straight Talk plan. We believe this is good news for MetroPCS and Leap Wireless, the two pure pre-paid carriers in the space.”
Quoting analysts from Current Analysis, FierceWireless said the transaction would put pressure on MetroPCS & Leap:
Interestingly, the analysts also point out that Sprint's acquisition of Virgin should hasten a merger between Leap and MetroPCS.
Whatever pride or bad blood has blocked the merger, such a merger is long overdue. The two have a roaming alliance and perhaps have buried the hatchet enough to form the nation’s 5th largest carrier (now that Alltel is gone).

The price wars in prepaid are only going to get more brutal, as MetroPCS and Leap both fuel the wars and respond to efforts by TracFone, Boost (Sprint), Virgin (Sprint) and GoPhone (ATT). Combining the two wouldn’t guarantee success, but it would expand and smooth their footprint while nearly doubling their respective scale for buying and other efficiencies.

Friday, June 19, 2009

Leap flying high

Leap Wireless has long been ignored by the industry, perhaps due to its market share, perhaps due to over expansion and Chapter 11 during the telecom bubble. More recently, it appeared doomed to be gobbled up by its rival Metro PCS, until they instead formed a roaming alliance.

On Wednesday, CNET ran an extensive Q&A with CEO Doug Hutcheson. Some of the answers are predictable, e.g. when other companies offer unlimited service it grows the market rather than hurting Leap (or its Cricket brand).

However, it offers a broad overview of how the prepaid carriers are both tapping unserved customers and also reflecting broader industry trends both here and elsewhere.

Perhaps most interesting are the profit numbers:
In some of our most mature markets, the average revenue per user (ARPU) is around $42 to $43. And we sell twice the number of voice minutes and two to three times the number of text messages as the large national carriers, which have ARPUs that are $10 to $15 higher than ours. And our profit margins fall in the high 30 percent range. In some older markets we are closer to 40 percent. We've said we expect to run in the mid-40 percent range.

Just to give you some perspective, these margins are a little lower than Verizon, but a little higher than Sprint. But we do it on much higher usage and lower ARPU. So this means that these providers can't expect to drop prices and lower ARPU and expect to make these same margins.
Leap has 90 million POPs — about 30% of the country, supplementing that with extensive roaming agreements (including that with Metro PCS).

Wednesday, May 13, 2009

Before Qualcomm

In the past few weeks, I’ve been hearing from old friends, my in-laws — and even a total stranger — about my article in the latest issue of the Journal of San Diego History.

The article, entitled “Before Qualcomm: Linkabit and the Origins of the San Diego Telecom Industry,” is the lead article of the Winter/Spring issue of the journal, which arrived in San Diego homes late last month. The text won’t be online for many months, and the journal doesn’t use abstracts. So let me quote the introductory paragraphs:
Most San Diegans know that our largest high-tech enterprise of the past decade is Qualcomm, a Fortune 500 company. Qualcomm—founded by Irwin Jacobs and Andrew Viterbi and five others—is recognized locally for the wealth that it created for founders, employees, and investors. It also enjoys a reputation for helping to fund many local philanthropic and charitable efforts.

Outside the local telecommunications industry, however, few people realize that the Qualcomm was the second company founded by Jacobs and Viterbi, not the first. The earlier startup, Linkabit, moved from Los Angeles to San Diego in 1971 and, over the next twenty-five years, helped to generate more than two hundred separate telecommunications companies. A series of fortunate coincidences—rather than a carefully thought out economic strategy—is the main reason why San Diego has a significant telecommunications industry.

This article traces the paths of Jacobs and Viterbi from their New England childhoods to their Massachusetts Institute of Technology (MIT) engineering degrees to the founding of Linkabit. It shows that Linkabit—its early application of cutting-edge communications technologies, its eventual sale, and the eventual dispersal of the company’s key engineering talent—was the catalyst that led to other new companies, including Qualcomm.
Because I expect to write other articles for the journal, I wanted my first JSDH article to tell the back story that few people know. Meanwhile, the journal wanted something with the “Q” word so that people would know why they were reading it. To achieve both goals, I worked with USD history professors Iris Engstrand and Molly McClain, who edit the journal on behalf of the San Diego Historical Society.

In the end, two-thirds of the article is about the story up until 1985 — when Jacobs and Viterbi left Linkabit, later starting Qualcomm a few months later. The final one-third is about the first 10 Linkabit spinoffs:
  • ComStream (1984)
  • Sciteq (1984)
  • Qualcomm (1985)
  • ViaSat (1986)
  • MultiSpectra Engineering (1986)
  • Pacific Communications Sciences, Inc. (PCSI) (1987)
  • Indra Technology (1987)
  • Primary Access (1989)
  • Milpower (1989)
  • Torrey Science (1990)
The final article was adapted from drafts of Chapter 5 and 6 of my (in progress) book, which cover Linkabit and its spinoffs. It also drew from my research into Andy Viterbi at JPL, which is both part of Chapter 3 of the book and my 2008 article on “Deep Space Communications as the Lead Market for Shannon Theory.”

Overall, we are all proud of the article. For me, it’s a chance to introduce myself to my hometown audience. For the JSDH, it’s a chance to get more local technology history in the journal.

What I really enjoyed was the chance to use so many photographs — something that’s never possible in an academic journal (except for a history journal). Most prominent were the color photos of Jacobs and Viterbi on the cover. Beyond these two men, I also got photos from Martha Dennis and some really cool ads from various archives. Together, they give a sense of what it was like for the Linkabit founders and employees back in the 1970s and 1980s.

The full article will eventually be online at the JSDH website. The historical society sells memberships (with subscriptions) at $40/year or back copies at $20 each.

Update, December 2009: The article is now available at the JSDH website.

Monday, April 27, 2009

Broadcom gets what it wants--but not 3G or 4G

On Sunday, Qualcomm and Broadcom announced settlement of their patent disputes in identical press releases from San Diego and Irvine. So far this morning, Qualcomm stock is up 6% and Broadcom stock is up 0.8%.

As with Qualcomm’s settlement last summer with Nokia, the agreement settles all existing disputes and lets Qualcomm and Broadcom get on with their business:
The agreement will result in the dismissal with prejudice of all litigation between the companies, including all patent infringement claims in the International Trade Commission and U.S. District Court in Santa Ana, as well as the withdrawal by Broadcom of its complaints to the European Commission and the Korea Fair Trade Commission. Under the agreement, the companies have granted certain rights to each other under their respective patent portfolios.
The press has emphasized the financial terms — Qualcomm paying $891 million over the next 4 years, including $200 million this quarter — which is clearly a lopsided settlement to allocate patent usage rights between the two companies.

TradingMarkets summarizes the numerous complex cases. However, the core issue was simple: Broadcom prefers royalty-free cross-licenses and has sought to coerce Qualcomm into providing same, using its various developed (and acquired) patents to win infringement judgements against Qualcomm.

Qualcomm sought at all cost (as it always does) to avoid setting a precedent that would undercut its core business model and revenue streams from the entire 3G (soon to be 4G) industry. It’s clear that Qualcomm paid dearly to settle the infringement claims without setting that dangerous precedent.

What seems to be under-reported is how the summary of the terms defines a clear partition of the industry:
  • Qualcomm’s customers get use of Broadcom’s patents for Qualcomm’s ICs in cellular products, but not for non-cellular products.
  • Broadcom’s customers get use of Qualcomm’s patents for Broadcom’s ICs in non-cellular products, but not for cellular products.
The only interpretation I have is that each firm is ceding key turf to the other: Qualcomm will stay focused (naturally) on cellular-related products, while Broadcom appears as though they will be avoiding cellular products that require Qualcomm’s patents.

At Qualcomm’s instigation, the first paragraph asserts
The terms of this agreement will not result in any change to Qualcomm's 3G (e.g., CDMA2000®, WCDMA and TD-SCDMA) and 4G (e.g., LTE and WiMAX) licensing revenue model.
The wording implies that Qualcomm changed its non-3G, non-4G revenue models. In other words, Broadcom got what it wanted — a royalty-free cross license — but not in cellular. This precedent seems to be the narrowest that Qualcomm could offer to satisfy Broadcom without hurting its core business.

On the analyst call, I would want to ask Qualcomm if the rights it granted Broadcom for non-cellular products will force it to renegotiate or otherwise change the terms of any other existing royalty agreements. My guess is that this is a small part of Qualcomm’s business, but as an analyst I wouldn’t want to guess.

However, the questions for Broadcom are more fundamental. It’s not clear what this means for Broadcom’s future, since it’s hard to see how they could ship a 3G or 4G product without a cellular patent license from Qualcomm. Had they already decided to give up on cellular and stay more in PCs and consumer electronics? Are they foreclosing cellular growth to avoid fighting Qualcomm? Or are they going to come back for a royalty-bearing agreement later on?

Saturday, April 4, 2009

San Diego's Android connection

Last week I was able to attend an Android event held in Sorrento Mesa. Despite the inextricable link between Android and its Google parent, San Diego has a strong representation in the Android ecosystem.

At the event — as at any gathering of SD telecom companies — the 600 lb. gorilla was Qualcomm. The big Q was one of the five founding members of the Open Handset Alliance.

Of the four presenters, two were local companies among the 34 OHA founders: Qualcomm and PacketVideo. Three other non-local founders were also represented: host Wind River (based near Oakland but acquired a San Diego operation) and attendees from LG and Sprint.

The panel was
  • Jason Bremner, Senior Director, Product Management, Qualcomm QCT (responsible for Multimedia, Winmobile, and all activities surrounding Android at QCT).
  • Dr. Cheuk Chan, Senior Vice President, CORE Client Products, PacketVideo (Dr. Chan's team is responsible for PV's OpenCore, which is the multimedia framework for Android)
  • David Tokunaga, Director of Technology Marketing, Nokia
  • Egil Gronstad, VP Technology Planning, Leap Wireless
Bremner talked how Qualcomm’s traditional two application platforms — Brew and Java — now include BlackBerry, Windows Mobile, Android and Linux. In the future it will also be supporting Brew MP, Flash and Symbian.

Qualcomm is providing chipsets for a wide range of Android devices. The first Android phone, the G1, was launched with Q’s MSM 7201A (which will also be used with the HTC Magic). Bremner said Qualcomm had the “industry’s broadest chipset support for Android,” with three levels:
  • QSD 8650: with 1 GHz MID, 720p HD video (suitable for Android-based netbooks)
  • MSM 7600: The replacement for the 7201A, a WVGA video chip suitable for smartphones.
  • MSM 7627: designed for sub $150 smartphones with VGA video
Chan talked about how formation of the Open Handset Alliance dates to 2006 — the year after Google acquired Android and a year before the public announcement of the Open Handset Alliance.

Tokunaga talked about general principles of open platforms in the context of the Nokia-owned Symbian open source platform. My ears perked up when he talked about “open enough” platforms — open enough to attract third party enhancements but closed enough to be controlled by a firm. (The reason my ears perked up is that “How open is open enough?” was the title of my 2003 paper on open source firms strategies).

Finally, Gronstad talked about this from Cricket’s standpoint. (Egil gave us a long intro on Leap, and several of us noted that the company and its Cricket brand are well known locally.) He talked about their plans to have its CDMA network open to all devices, both those it sells and those it does not. Someday (perhaps not soon) there will be a Cricket smartphone — just as there is already a Metro PCS smartphone (a two-year-old Qualcomm-enabled BlackBerry 8830 once exclusive to Verizon).

Qualcomm working with vendors for more than 20 Android-enable devices. However, a third party developer noted that handset vendors are holding off on releasing new phones because the performance and user experience is not there. Bremner said that “On the G1, we compromised on featureset and performance” to get the device to market, but now the top priority at the Q is to optimize Android performance with the MSM chipsets.

Overall, this partly answers one of my long-term questions. With CDMA going away as a separate air interface, what will be the Qualcomm-specific ecosystem or the benefit of wireless companies to locate in San Diego? There clearly is an increasing role for QCOM chipsets in shipping phones on the CDMA side, and also for Q’s expertise on platform support, thus giving local application developers a leg up over most of the rest of the country.

Thursday, March 26, 2009

Dr. D helps take over Wind River

This morning at the (impossible to find) Wind River offices on Scranton, Marco Thompson hosted the “Android VIP Roundtable III” with four expert panelists on the Google-sponsored open source platform.

As part of hosting the event, Russ Christensen of Wind River put in a plug for the company’s hardware and software design activities for mobile — including Android-based devices. It struck me how much Dr. D has helped transform Wind River.

Once upon a time, Wind River was an embedded tools and OS (VxWorks ) licensing company that aspired to be the Microsoft of the embedded world. They spent a lot of time predicting such an outcome — I remember one Embedded Systems Conference where the pitch was particularly strong — but the industry for years remained badly fragmented.

Now the industry is consolidating around a common platform, but it’s not VxWorks. Instead, it’s Linux, which has many other suppliers beyond Wind River. And the common toolset for the embedded world is based on Eclipse.

So with the proprietary licensing business shot, Wind River has found a new niche around being a solutions provider. The pitch today on Christensen’s slides emphasized time to market.

This was always the Dr. Design story. Founded in 1984 by Marco (UCSD ’79), their speciality was to solve seemingly impossible problems in an insanely short period of time (and charge accordingly).

In 1999, Wind River paid $400+ million to acquire solutions provider Integrated Systems, Inc. In 1996, ISI had paid about $17.5 millionto buy Dr. D.

So what once might have looked like a less-scalable, lower margin complementary business to Wind River’s platform sale has now become an integral part to their new business model based on integrated solutions that integrate an arbitrary combination of hardware and software, off-the-shelf and custom components. This is exactly what Dr. D did for 12 years before Marco cashed out.

The results are not exactly as dramatic as the NeXT takeover of Apple that ushered in the Jobs II era, but in this era of commoditized software platforms, it’s still a happy outcome as Wind River seems to have made the transition to fight another day.

Wednesday, March 25, 2009

Cricket's big idea

Giant-Samsung-Messager-PhoneIn Philadephia, Cricket and Samsung are unveiling a 13' high replica of the Samsung Messenger.

The two companies earlier showed the giant phone in Chicago, as part of a publicity stunt for Cricket’s entry into the local market.

San Diego’s Leap Wireless (which owns the Cricket brand) has expanded to 35 of the top 50 markets with a POP of about 84 million — or about 27.5% of the potential US market. At the end of 2008, it had about 3.8 million subscribers.

Its younger roaming partner/low cost rival/spurned suitor Metro PCS is covering both unique and overlapping markets, with coverage in 92 of the top 100 markets. It ended 2008 with about 5.4 million subscribers.

Together, the two carriers have about 3.5% of (my estimate of) the 265 million US cellphone subscribers at the end of 2008, far behind the 28% of #1 Verizon Wireless (plus Alltel) or even the 12.4% of the smallest of the surviving national carriers, T-Mobile.


Photo from Chicago taken from Techgadgets.in.

Tuesday, March 24, 2009

Nokia joins The Network

After largely giving up on CDMA in the Americas, on Monday Nokia introduced a stylish new flip phone, the Nokia Intrigue 7205.

The new phone is being offered exclusively on the largest US carrier, CDMA-based Verizon Wireless. It is the third CDMA phone here in the last year for Nokia, beginning with the 6205 and more recently the 2605.

It is unclear whether Nokia will have any impact on the LG and Samsung dominance of featurephone sales at Verizon stores across the country, but it is the logical next step after the patent settlement with Qualcomm.

Sunday, March 22, 2009

Broadcom misfires -- this time

When Broadcom filed suit against Qualcomm last October, it was the latest attempt by the Orange County low cost chip producer to force Qualcomm to change its business model.

Qualcomm announced last week that the suit was dismissed by the US District Court. However, if I were still a Qualcomm shareholder, I’d still be a little worried.

The suit was an opportunistic effort to use the SCOTUS decision in Quanta v. LG as a cudgel. As USD Law Prof. David McGowan analyzed the Broadcom complaint in this blog last November
The complaint alleges Qualcomm double-dips by charging handset manufacturers a royalty on both chipsets and handsets. Qualcomm may earn chipset royalties through sales to handset manufacturers or from licenses to competing chipset manufacturers such as Broadcom. It also charges a royalty directly on handsets.

Broadcom claims the Quanta case holds that a patentee may not charge downstream royalties beyond the first sale of its technology. This claim implies Qualcomm may either charge handset manufacturers a royalty for the handset or for a chipset (directly from Qualcomm or indirectly from Broadcom) but not both. Broadcom is asking the court to declare that it need not pay royalties to Qualcomm when Broadcom sells chipsets to handset manufacturers who pay Qualcomm royalties on handsets.
Even for a law professor (who are generally cautious about predicting court outcomes), McGowan saw the case as too close to call:
It is hard to say what is likely to happen. Quanta dealt with an unconditional license and Qualcomm has said its licenses are expressly conditional. The Court did not rule on such licenses. The district court therefore is not bound to rule for Broadcom. If one read the Supreme Court as sending the signal that patentees should only be able to engage in one transaction per product, though, one could extend Quanta to reach the result Broadcom seeks.
There isn’t much coverage of the ruling online, of which the Telephony Online explanation is the most incomplete:
US District Court Judge William Hayes, however, rejected Broadcom’s reasoning, ruling that Broadcom provided no specific of exhausted patents and that the purported damages Qualcomm has wrecked upon the industry were too speculative.
So I’m no law professor and in fact never went to law school, but the case was clearly dismissed on procedural grounds, not on a question of law; Broadcom’s legal claim was never tested. To me this suggests that the Broadcom case was rushed to court, without bothering to document specific patents and to find a putative victim. Obviously we don’t know what happens within Broadcom’s legal team, but this seems like the sort of weakness that a first year associate could have spotted before the case was filed.

I don’t know the specifics of the ruling, so it’s not clear whether the complaint was dismissed with prejudice, and if so, whether it prevents Broadcom from ever filing another patent exhaustion claim again. Even if so, one of Qualcomm other rivals might want to pick up the cudgel.

Telephony Online noted that Nokia filed a similar complaint that was thrown out (but that was pre-Quanta.) Nokia and Qualcomm now claim to be friends, so perhaps for now Nokia will train its legal guns on InterDigital rather than Qualcomm.

But Qualcomm’s business model and market share leave no shortage of enemies. If not Broadcom or Nokia, another possibility is TI — which was the world’s largest supplier of wireless communications chips until Qualcomm passed it last year. Finally, there are the Koreans, who never stop complaining about CDMA royalties and (like Broadcom) have been looking for any possibility to reduce the checks they send to San Diego.

Thursday, February 19, 2009

QCOM: technical vs. fundamentals

Normally I don’t write about Qualcomm’s stock, but Qualcomm’s stock value got about the most prominent discussion possible on Thursday.

On his nightly CNBC show, retail stock pundit Jim Cramer spent several minutes Thursday night analyzing the last few months of momentum on Qualcomm’s stock, in his weekly segment called “Sell Block.” A written summary is provided on the Mad Money website (and also in a separate article).

He started by talking about Qualcomm’s fundamental business, including a simplified version of its business providing IPR and chips for cellphones. Cramer said: ”It reminds me of the old Intel inside ads. It should be Qualcomm inside.“

But the impetus for the discussion was a very negative report from technical analyst Helene Meisler. To briefly summarize
  • Any rise in Qualcomm’s stock price is blocked by its January peak.
  • The upward trends that started earlier this year has stalled.
  • When there is higher stock volume, the trend on those high volume days is “telling the truth,” and Qualcomm had some ugly down high-volume days.
  • The stock is heading down towards $30 and won’t break $40 again. (The stock closed at $33.84 Thursday).
Cramer, on the other hand, argues that the business fundamentals favor QCOM. It has been taking share against its rivals. It earns $4-$8/phone royalties (average phone price $200), with 90% margin. Worldwide 3G penetration to go from 40% to 70-80% over next 3 years. This week, Qualcomm even won cooperation from former archenemy Nokia.

Cramer was also not surprised by the lower CY 2009 earnings guidance that Qualcomm provided with last month’s earnings release. To Cramer (and to me), there is the implication that Qualcomm assumes that the economic recovery is pushed back at least to 2010 — a realistic assumption.

Cramer thinks it’s a good deal because it’s down from its $56 peak (last August 14). As Cramer said, “A lot of the bad news is baked in” praising its “pristine balance sheet.”

I don’t know which one is right, but I know they both can’t be right. If over the next 6 (or 12) months, the stock never breaks $40 and ends below $30, clearly Meisler. If it ends above $40, Cramer is right. If it just muddles along between $30 and $40, Meisler isn’t wrong, but that’s a pretty weak sell call — particularly if (as I expect) the broader market is sideways or down for the next 18-24 months.

Note: Cramer owns QCOM for his charitable trust. I do not own any shares.

Tuesday, February 17, 2009

Qualcomm's other shoe

After 10 years in which Symbian never supported CDMA or any Qualcomm chipsets, why did Qualcomm join the Symbian Foundation last week? As the EE Times reports, the other shoe dropped this morning:
BARCELONA — Both ST-Ericsson and Qualcomm Inc. have revealed partnership programmes with Nokia based round reference platforms that will use the Symbian Foundation's software.

Qualcomm's deal focuses on developing UMTS mobile devices, initially for North America, that will be based on the S60 software on Symbian OS, running on the San Diego, California- based chip maker's latest MSM7xxx-series and MSM8xxx-series chipsets targeting wireless broadband.

The companies, for long arch rivals due to long standing patent infringement law suits that were finally settled last year, say the first mobile devices based on the collaboration would be expected to launch in mid-2010 and be compatible with the forthcoming Symbian Foundation platform.
Nokia currently has a very limited relationship with US carriers. “UMTS … for North America” means selling S60 smartphones either to AT&T or T-Mobile, competing for shelf space with the iPhone or gPhone respectively. Certainly, outside the US the Nokia smartphones are sold side-by-side with the iPhone by the same carriers.

EE Times plausibly argues that this cooperation is because Nokia last summer finally resolved its patent fight with Qualcomm. In November, analysts told EE Times they thought it unlikely that Nokia would go so far as to buy chips from Qualcomm.

It seems particularly odd that Nokia would start its partnership with Qualcomm with W-CDMA phones — given its sells hundreds of millions worldwide — but perhaps it is just trying to prime the pump. Presumably the next step is to use QCOM chips to sell more CDMA phones in the US, although so far Nokia is outsourcing that work to ODMs.

On a somewhat related note, Qualcomm CEO Paul Jacobs was quoted in Barcelona by CNBC as saying demand for smartphones remains strong. However, unlike his predecessor (Jacobs père), Jacobs fils is a perpetual optimist.

TI and InterDigital ally against Qualcomm

In connection with GSM Mobile World Congress, TI and InterDigital have made announcements reflecting an alliance between them. The new development is helping InterDigital make its transition from being an IP licensing company (that a few rivals have called a patent troll) to a firm that makes 3G chips to subsume its IP.

Not surprisingly, InterDigital is more enthusiastic about the alliance than TI is. (It’s fun to read these tea leaves — as someone who ran a little company that did joint press releases with big companies — the message is pretty clear). InterDigital issued a press release trumpeting its W-CDMA (i.e. HSPA) modem option and how it works with the the TI OMAP 3 and OMAP 4 processors:
"We are pleased to be the wireless modem supplier for TI's advanced OMAP 3 platform. Our high performance HSPA modem offers instant mobile broadband connectivity, accelerating the development of compelling new applications," stated Mark Lemmo, Executive Vice President, Business Development for InterDigital. "Available as a 3G modem option, InterDigital's SlimChip MID Module has been pre-integrated with the Zoom OMAP34x-II MDP, allowing mobile application developers and OEMs to fully exploit the rich capabilities of this platform."
The press release goes on to quote Bill Crean, Strategic Marketing Manager, as praising “ The flexible architecture of the OMAP 3 platform allows it to easily connect to InterDigital's SlimChip MID Module.”

Meanwhile, the TI press release allows InterDigital space with other vendors who praise themselves and the OMAP 3. The only mention of InterDigital comes from InterDigital itself:
"We are pleased to be a wireless modem supplier for TI's advanced Zoom II mobile development platform. Our high performance HSPA modem offers instant mobile broadband connectivity, accelerating the development of compelling new applications," stated Mark Lemmo, Executive Vice President, Business Development for InterDigital. "Available as a 3G modem option, InterDigital's SlimChip MID Module has been pre-integrated with the OMAP34x-II MDP, allowing mobile application developers to fully exploit the rich capabilities of this platform."
Eleven months ago, blogger Vijay Nagarajan predicted on his own blog (and later on Seeking Alpha) that an alliance with InterDigital would help TI:
TI is the market leader for 3G application processors with its OMAP product-line. But it does not currently have a standard 3G baseband solution. The company’s OMAP roadmap merely has placeholders for future merchant ICs with 3G baseband. Its mammoth market share comes from custom chips it develops for Nokia and Motorola among others. This position is, however, challenged by the multiple sourcing trend that the handset vendors are now adopting. Left behind in the 3G baseband race, not only by Qualcomm but also by Broadcom, Infineon, InterDigital etc, TI is finding itself losing mobile share. The aggressive strategy and the niche product positioning by the competition is not helping its cause either.
So although Qualcomm has a unique combination of technologies, its rivals are able to ally to produce a competing combination. (We call this “open innovation.”) I don’t know how this will impact Qualcomm’s profitability in the long term, but it seems as though the increased competition isn’t going to help.

Monday, February 16, 2009

Qualcomm joins Symbian Foundation

Cross-posted from OpenITStrategies

In advance of this week’s GSM Mobile World Congress, Qualcomm has joined the Symbian Foundation (as have HP and MySpace). Qualcomm of course is interested in supporting CDMA around the world, in addition to its dual mode 3G processors.

Rumors of Symbian CDMA support date to at least 2004. Nokia had once sponsored an adaptation of Symbian S60 to work with CDMA – presumably to gain access to the US market and segment where it’s had a relatively weak presence.

Back in 2005, MobileTracker reported that Nokia had won FCC approval for the Nokia 6638:
The FCC has approved (FCC ID QMNRM-18) the Nokia 6638, the first CDMA Series 60 phone. Visually, it looks just like the Nokia 6630 with the addition of a [very long] antenna. Since the 6638 is a CDMA handset, Nokia is most likely aiming for a US release.
The phone was compatible with Verizon’s two CMDA bands, 800 MHz and 1.9 GHz.But the phone apparently never made it to the market.

Qualcomm has already joined the Android Foundation. Apparently Qualcomm has worked on a Windows Mobile prototype that works on Qualcomm chips, and its chips power several BlackBerry models that are dual cdma2000/W-CDMA, including the Storm and the 8830 world phone.

Now Qualcomm has committed resources to the major smartphone operating systems — pretty much everything outside the iPhone. From now on, we’ll see how many operators and manufacturers use Qualcomm chips for their Android, Symbian or Windows Mobile smartphones.

Friday, February 13, 2009

Korea 3G finally shifts to W-CDMA

For 2G, Korea was exclusively cdmaOne, the only country in the world where this was true. Operators quickly got into 3G by offering cdma2000 on their existing frequencies, but the government licenses specified W-CDMA for new 3G frequencies that (IIRC) were compatible with the Japanese selections. As with anything else in Korea, each round of choices was designed to protect (or at least help) the export efforts of Korean companies.

The website Cellular News reported Thursday that in Q4 2008, the total W-CDMA subscribers in Korea finally passed those for cmda2000: 16.5 million vs. 14.9 million. For me, this marks the end of an era for Qualcomm’s original success story in Korea, marred only by the periodic royalty complaints.

Thursday, February 12, 2009

CDMA success brings Chinese royalty gripes

Taking a page from their Korean counterparts, in the face of rapidly growing sales, Chinese cellphone manufacturers are airing complaints through local propaganda organs in hopes of lowering their contractually obligated CDMA royalties.

Last year, a reorg of China’s telecommunications industry brought dominant landline provider China Telecom the opportunity to offer 3G mobile services using Qualcomm’s cdma2000. On Wednesday, China’s official Xinhua news service reported good news for Qualcomm and its local licensees:
Under the stimulation of China Telecom's large investment and procurement this year, the growth of China's CDMA cell phone market will be great, said an industrial insider. Competition will heat up as many GSM mobile phone manufacturers are entering the CDMA field, he added.

China's CMDA cell phone market rebounded in the end of 2008 despite gloomy market conditions. According to statistics released by market research firm Sino-MR, sales volume of CDMA handsets topped 1.29 million during December 2008, up 33.6 percent year on year and 183 percent month on month, marking a five-year high.
However, the report ended with a complaint about the royalty rates
Insiders with China Telecom consider China an important market for CDMA business and expect Qualcomm to realize the market condition and give more supportive pricing. Qualcomm refused to comment.
When in 2000 Irwin Jacobs negotiated government permission for CDMA to be used in China, Qualcomm granted Chinese manufacturers the best rates in the world — around 2.65% as opposed to 5% for other makers. (Qualcomm neither discloses nor confirms its royalty rates).

In fact, this deal has brought continued griping by Korean manufacturers paying list price. The Korea Times has run one-sided stories since 2004 airing selective leaking by the unhappy manufacturers. For example, it ran a March 2004 story (no longer on their website) entitled “Qualcomm's Royalty Policy Angers Korean Chip Makers”:
U.S. chipmaker Qualcomm’s royalty policy has angered Korean manufacturers of code division multiple access (CDMA) phones as it charges lower rates to China than to Korea.

A local CDMA phone maker contends that Qualcomm gives a “subsidy” to Chinese exporters as it collects as low as 5 percent of handset prices in royalties from them. But the rate is 5.75 percent for Korean makers.

Korean cell phone producers, which pay a fixed 5.75 percent for export and 5.25 percent for local sales to the San Diego-headquartered firm, shoulder higher royalty burdens in both domestic and global markets.

According to a contract between Qualcomm and one of China’s leading handset manufacturers, which The Korea Times obtained exclusively, the royalty rates are 2.65 percent for local sales and 7 percent for offshore shipments.

However, the contract says the export royalty will be cut to as low as 5 percent after the three years immediately following the license effective date.

The deduction ranges from 5 percent to 6.5 percent, depending on quarterly export volume, but the Chinese firm is entitled to enjoy the low rates “if more than 100,000 such subscriber units (handsets) are sold in the applicable calendar (year)” in overseas markets.
The KT report said that Chinese manufacturers signed a 10-year agreement in “early 2001,” which would mean these deals will be expiring in 2011.

A review of Qualcomm’s 2001 press releases reveals this May 21, 2001 announcement:
Qualcomm Incorporated … today announced that it has signed a commercial license with ZTE Corporation, a leading manufacturer of telecommunications equipment in China. Under the terms of the royalty-bearing license agreement, Qualcomm has granted ZTE a license under Qualcomm's CDMA patent portfolio to develop, manufacture and sell cdmaOne™ and third-generation (3G) CDMA2000 1x/1xEV network equipment. The license grants ZTE the right to use Qualcomm's technology and integrated circuits to make and sell cdmaOne and CDMA2000 equipment in China and worldwide. ZTE becomes the first company in the People's Republic of China to enter into a commercial license with Qualcomm.
Additional manufacturers signed licenses in July and November, with the latter including Huawei. Another 17 Chinese firms signed in January 2002.

The Xinhua story seems designed to help Chinese manufacturers (or China Telecom) pressure Qualcomm to lower royalties on the current deals, or at least when the licenses come up for renewal in two years.

Interestingly, there is no complaining about the GSM royalty rate, which is also secret. One report estimated the rate at 2-10%. However, data from the various Sendo-related lawsuits put the number more like 10-13%, suggesting that single-digit GSM royalties are only available to firms with big enough patent portfolios to negotiate a cross-license. Thus far, only a few Chinese manufacturers have enough patents to possibly put them in the latter category.

Thursday, February 5, 2009

MediaFlo hurt by DTV delay

The House voted yesterday to delay the Digital TV transition by nearly 4 months. A majority of both houses have decided that America is not ready for a transition and thus a delay to June 12 is warranted.

However, delaying the end of analog transmissions (and thus the surrender of those broadcast bands) is also delaying the deployment of new services by the communications companies who paid for rights to those bands, especially in the various 700 MHz bands.

Although AT&T and Verizon own such spectrum, the company most ready to use its spectrum is Qualcomm with its Media Flo service licensed by AT&T and Verizon.

Qualcomm issued a statement Wednesday (although it’s not on its website), with the most complete summary in Telephony magazine:
We are disappointed with the passage of legislation extending the DTV transition date to June 12th. … Due to the investments we made, we were ready for a February 17th transition to provide our innovative FLO TV service nationwide immediately. We are encouraged that several Congressmen and Senators who supported the delay stated that this would be a one-time delay only. In light of the fact that the legislation, as amended and finally passed by Congress, allows TV stations to transition voluntarily between now and June 12th, we cannot determine the specific impact of the final bill's passage on our MediaFLO business.
Due to interference its service causes around (analog) UHF channel 55, Qualcomm will have to delay rollouts in Boston, Houston, Miami and San Francisco; its efforts to have these markets exempted fell on deaf ears.

It’s hard to predict the commercial impact of the delay. Qualcomm has been aggressively planning this rollout for five years, and its speed to market was intended to preempt rivals (at least in the US) from gaining a foothold in the (presumed) lucrative mobile multicast TV market.

Sunday, February 1, 2009

Coding in space

Local telecom industry veterans, particularly those who got a good dose of information theory in their EE degrees, understand why Andy Viterbi won his Shannon Award. All would know that Viterbi developed what (proved to be) the maximum likelihood decoding scheme for convolutional codes, which remained the state of the art until turbo coding came along.

Most would know that every GSM cellphone (let alone CDMA cellphones) includes a Viterbi decoder. Some might know that one of Qualcomm’s first products (and I believe its first semiconductor product) was a Viterbi decoder chip.

A few would recall that Klein Gilhousen led the group that designed the LMP (Linkabit Microprocessor) which was used to implement the Viterbi algorithm for the Dual Modem.

What might not be known — I didn’t know until I researched the back story for the book — is that before cellphones, chips and the dual modem, the first real application for the Viterbi decoder was for NASA deep space probes.

In the late 1960s, NASA was planning two competing families of space probes for mankind’s first visit to the outer planets. The communication challenges were among the most daunting in history, where (due to the inverse square law) the signal strength a billion miles away was faint, and power (from nuclear decay batteries rather than solar power) is minimal.

Theoretically, the application of Claude Shannon’s ideas (and coding theory) could provide as much as a 10 dB signal gain. MIT-trained academics worked with two different divisions of NASA, both using coding theory:
  • Pioneer 10 and 11: Codex of Newton, Mass. worked with NASA Ames to apply Fano decoding
  • Voyager 1 and 2: Andrew Viterbi, Joe Odenwalder and others — first as a UCLA contract and then as a Linkabit contract — worked with JPL to use concatenated decoding and the Viterbi algorithm.
My research on this led to chapters on MIT and NASA for the book, as well as changes to other chapters. However, I’ve stopped quoting an ETA for the book given (as a part-time job) it’s coming very slowly.
In the meantime, two years ago I wrote a paper for the Journal of Management Studies, in a special issue on technology commercialization. The paper, entitled “Commercializing Open Science: Deep Space Communications as the Lead Market for Shannon Theory, 1960–73,” was published in December after several rounds of review.

Errata (Feb. 4, 2009). I received an email from Andrew Viterbi with a correction to the paper that will be applied to the book:
the first Linkabit customer was the Naval Electronics Laboratory Center (NELC) in San Diego… NASA was the second, but never became a large customer.
Most universities subscribe to the journal, so the paper is available as online from a university IP address (DOI: 10.1111/j.1467-6486.2008.00807.x). Normally I post a PDF of the pre-publication Word document, but there were major changes after submission during editing the page proofs. However, I would be glad to email the published PDF to anyone who asks (at ).

Graphic credit: Brewster Rocket comic from Sunday, February 1, 2009. Note to Tim Rickard: Pioneer 10 was launched first, but Voyager 1 was actually the first manmade object to leave the solar system and the most distant space probe. So your aliens would first see the Golden Record rather than the Carl Sagan memorial plaque.

Tuesday, January 20, 2009

Qualcomm expanding CPU reach

Qualcomm today spent $65 million to buy what’s left of AMD’s handheld devices division. The division, which made components for consumer electronics, was part of AMD’s 2006 acquisition of ATI. The acquisition included some technologies that Qualcomm previously licensed from AMD.

Apparently AMD plans to stay in the netbook market, where it competes with Intel’s Atom and also Qualcomm (which sells wireless chips today and promises Snapdragon someday). AMD continues to be struggling after its ATI acquisition, which largely failed.

Monday, January 12, 2009

Viral Tell Zone now largest US carrier

Verizon has completed its acquisition of Alltel, making the CDMA carrier the nation’s largest, with 83.7 million subscribers and 28% market share. Verizon promised
Since Verizon Wireless and Alltel use the same technology platform, the vast majority of customers will be able to use their current handset after the transition to Verizon Wireless.
About 2.1 million subscribers across 105 (mostly rural) markes in 24 statets will be divested to meet antitrust concerns, with Verizon dumping a mix of Alltel, Verizon and Unicel (acquired last August) branded properties.

Verizon said it will integrate Alltel name, products, services and distribution in Q2. The LAT spotted a missed opportunity:
Sadly, that probably means slapping Verizon names on the Alltel properties, rather than some strange combination such as AllVerizon, Veritel or Allizon. Too bad they don't want to use an anagram of Verizon Alltel: the best one we could find was Viral Tell Zone, which seems fitting for a wireless company.