Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Strategies & Market Trends : Gorilla and King Portfolio Candidates -- Ignore unavailable to you. Want to Upgrade?

To: Mike Buckley who wrote (36416)	12/11/2000 10:35:38 PM
From: Apollo	Read Replies (1) \| Respond to of 54805

Intel....more from Ten in response to Merlin: If s/he had clarified in detail the basis of the opinion that "the x86 architecture is now an open architecture that Intel can no longer keep to itself like Microsoft can with Windows," my vote might have been swayed. That's because I assume Ten is saying that the architecture is no longer essentially proprietary. Is that because critical patents have now expired or is it because of Intel's reaction to market conditions? I brought this question to Ten; Ten's answer: "Apollo, I have very little knowledge about x86 licensing and all that. I do remember that there was a spat between AMD and Intel over AMD's act of copying MMX, an x86 extension that Intel developed. After some legal scuffles, I believe they settled with AMD agreeing to some sort of licensing deal. After that, AMD decided to go ahead on a completely different path than the one Intel is taking. AMD introduced 3DNow extensions in the middle of 1998, hoping to get game developers before Intel introduced SSE in early 1999. However, SSE has won out, because of technical superiority over 3DNow, Intel's huge market share, and Intel's own efforts in software development (especially Intel's own compiler package which is fully supported and marketed by Intel). And now Intel is pushing further with SSE2 on the Pentium 4. Now AMD is going their own route with x86-64, their version of 64-bit extensions for x86. This is in marked contrast to Intel, who is developing Itanium pretty much as a brand new computing platform. However, AMD also announced support for Intel's SSE and SSE2 in their first x86-64 processors. How AMD can implement this without Intel's blessing, I don't know. However, I can't imagine Intel would let this go by without a legal fight. But I do know that neither SSE, nor SSE2, nor even 3DNow are considered "must-have" features in an x86 processor. I don't know of any software package which explicitly requires support for one of these extensions. It's because the software, drivers, and operating system can all work around the lack of these extensions. For example, there are many device drivers and software applications which will work better on an SSE processor, but all of them will still work on non-SSE processors. So in essense, these enhancements are only for enhancing performance, and only for certain software packages. Unfortunately, these days, performance isn't as much of a selling factor as it used to be. This, plus the ability to work around the lack of such extensions, is why none of these extensions are seen as "must-have." Tenchusatsu"

To: Mike Buckley who wrote (36416)	12/12/2000 11:42:57 AM
From: Eric L	Read Replies (2) \| Respond to of 54805

Mike, Re: Intel & Qualcomm - Gorillas or Kings? << Regardless, if Ten is correct about that, it presumes that a Gorilla market can morph into a King market. Do we have any examples of that happening in the past? >> The recent discussions about Intel (which I find very worthwhile although I am rather adamant about Intel's gorillahood), lead me to question whether Qualcomm is destined to morph into a king. The issue is not having simply essential IP in a technology, it is IMO, having architectural control of the technology in either de facto or de jure manner. The initial CDMA tornado is over. Hypergrowth has subsided rather dramatically. Qualcomm is now enjoying CDMA main street. We will not see >100% YOY subscriber growth of CDMA again for some time, and it is doubtful we will see it again for the architecture with high switching costs that Qualcomm has proprietary control over. All indicators are that in the second half of this decade W-CDMA will become the dominant air interface for mobile wireless providers. While Qualcomm has essential IP in W-CDMA, they have no architectural control over it. W-CDMA is a committee controlled architecture. Qualcomm does not participate in the committee (3GPP) that has been standardizing the architecture of the air interface and integrating it with the network subsystems(s) since early 1998. Qualcomm fought vigorously for a single converged standard. Dr. Jacobs commented openly on this in a letter to Business Week in September 1998 here: Message 5849466 Instead a compromise was reached and rather than a single converged standard we have multiple harmonized standards. In 1998 Qualcomm published a white paper called 5 Points: Message 6207324 3GPP W-CDMA recognizes none of the first 3 principles of Qualcomm's 5 Points. Qualcomm is in the process of harmonizing to 3GPP W-CDMA in 3GPP2 rather than the other way around, while concurrently evolving their own cdma2000 standard. In addition, even within 3GPP2, Qualcomm is now more involved with the evolution of a committee based architecture, than they ever have been in the past as it relates to evolution of 1xEV-DV. Moreover, so long as Qualcomm stays resident in 3GPP2, they run the risk of further losing the capability of influencing the architectural of W-CDMA, IMO. For the moment, however, I still categorize both Intel & Qualcomm as gorillas. One of these days I intend to go back and review Ferguson & Morris's "Computer Wars" Moore states "were the first to explicitly point out the importance of architectural control to market-place power" and which Moore freely admits to appropriating" when defining "Proprietary Open Architecture with High Switching Costs (page 52; RFM). Interesting topic. - Eric -