Eric,
For VERY valid reasons Qualcomm is getting dragged into our discussion today of Intel and qualifications for gorillahood in the chip business.
<< The difference is that Qualcomm, in essence, is an IPR company, and depends on the royalties paid by the companies it licenses its technology to for its income. >>
That is part of their revenue stream, but not all. Note that I am talking about Qualcomm today, not Qualcomm tomorrow (after the SpinCo spin, it being darned hard to keep up with Qualcomm's business model) and not yesterday when they competed with their licensees and value chain in infrastructure and handsets, but were already a gorilla as determined by the consensus of these gorilla hunters, if not by Geoff.
<< On the other hand, not only does Intel depend on the sale of its products for income, it gains nothing by having given AMD a license. >>
You might be correct if we were discussing conventional investment theory.
According to Moore, however, "Other companies leverage the hypergrowth market of the gorilla ... IBM insisted that Intel license its 8086 chip architecture to second sources ... and this proved a boon to Intel's own growth. The market grew incredibly fast, thereby institutionalizing Intel's architecture as a de facto standard."
This thinking parallels that of Ferguson and Morris ("Computer Wars") who initially defined concepts refined by Moore as "Proprietary Open Architecture", and packaged as the key defining characteristic of a gorilla when initial hypergrowth in a market sector commences.
<< Qualcomm's licensees are its customers, while Intel's "licensees" are its competitors. >>
Perhaps. But Qualcomm's licensees have been and in many cases still are its competitors. This in itself presents some potential conflict and animosity.
Your comments, and those of others, encouraged me to go back and reread the cover story from the March 2000 Issue of Cahners "Electronic Business" Magazine called "Meet the new Qualcomm" that addresses the issues and challenges that Qualcomm faces in the IC arena. I've clipped one applicable section from the "Pre SpinCo" article below, and highlighted some of the excerpts from the article which interested me when I first read it, but also bear on the discussion in progress.
>> COUNTING ON CHIPS
eb-mag.com
That sort of rapid growth will be crucial to support Qualcomm's transition from phone maker to designer of chips and software. To its credit, the company already has shipped more than 100 million CDMA chips. While most competitors are just reaching the market with their first CDMA devices, Qualcomm has already begun shipping samples of its seventh-generation MSM5000 CDMA chipset, a transitional step toward the full CDMA 2000 standard and the first available 3G technology.
Qualcomm relies on Taiwan Semiconductor Manufacturing Co.,IBM Corp., Intel, Texas Instruments Inc. and others to fabricate its chip designs. Last year, Qualcomm vaulted past San Jose, CA-based Altera Corp. to become the world's largest fabless chip company. Being fabless keeps it nimble and saves it the huge cost of building manufacturing plants. Yet as CDMA volumes rise, integrated chip makers like Intel and LSI Logic may gain advantages by deriving benefits from their manufacturing expertise.
Another newcomer with the potential to become a heavyweight is Eindhoven, Netherlands-based Philips Electronics NV. Last year, it bought CDMA chipset maker VLSI Technology Inc., and has been haggling with Qualcomm since then over the transfer of VLSI's CDMA technology license from Qualcomm. Motorola, an early CDMA licensee, could decide to begin selling its chips to outside customers, although thus far it has not met even its own internal needs. One company that doesn't look like a competitor is Kyocera, which has agreed to continue buying Qualcomm's CDMA chips for at least five years.
Qualcomm had little choice but to make its own CDMA chips, and license its technology to other chip makers, until the market took off. But it's not clear whether it can continue doing both over the long term.
Will Strauss, president of Forward Concepts Co., a Tempe, AZ, market research firm, sees inherent conflicts between Qualcomm's own chip business and its CDMA licensing. "They're competing with their ASIC licensees," he says. Strauss sees Qualcomm's business strategy as not quite like Intel, which tries to dominate production of a key technology component, nor like that of U.K.-based ARM Holdings PLC, a "chipless" semiconductor company that derives revenue almost totally from technology licensing. "There are no models for what they're doing, none whatsoever," he says.
Jack Quinn, president of Micrologic Research, says he expects Qualcomm to eventually get out of the CDMA chip business and rely mostly on license and royalty revenue. "Right now there's a good markup in [CDMA chips], but it won't always be that way," he says. "That chip business is going to be dog-eat-dog."
Qualcomm's Jacobs concedes his company has created numerous competitors for its CDMA chip business. Yet he has every intention of keeping Qualcomm's chip business out in front. "We've got a very good lead as far as CDMA is concerned and we're working hard to extend that lead. . . I don't see us selling that business." <<
Good discussion, courtesy of saukriver. Thanks for joining in.
Regards,
- (other) Eric - |
