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.

Technology Stocks : Intel Corporation (INTC) -- Ignore unavailable to you. Want to Upgrade?

To: Patient Engineer who wrote (44027)	1/5/1998 8:27:00 PM
From: Mo Chips	Read Replies (2) \| Respond to of 186894

Uh oh, your going to get it from Paul... I do, by the way, agree with your concerns. Very well said. I am thinking long and hard about dumping my shares if it hits 80. Its been a long good ride, but there seems to be limited upside in my opinion, for the next 12 months or so... Mo

To: Patient Engineer who wrote (44027)	1/5/1998 9:19:00 PM
From: Paul Engel	Read Replies (1) \| Respond to of 186894

Patient Engineer - Re: " I have to agree with most of what you say about Intel. They are unbeatable as a microprocessor vendor....So how does Intel get EPS growing again?...I suspect you disagree. I would love to hear your reasoning." Thanks for your thoughts - it's been a long time since I heard from you! Your concerns are valid but they miss some obvious points. You assume that ASP's will continue to drop. According to Ashok Kumar (who may be blowing smoke) Intel actually RAISED ASP's in Q4 from $215 to $240: {===================} news.com ""We forecast microprocessor units to have grown about 5 percent sequentially and 14 percent from the year-ago period to 22.8 million. Microprocessor ASPs are expected to have recovered from the depressed level of $215 in the third quarter to about $240 in the December quarter." {===========================} Now how did they do this? High performance Notebook CPUs (Tillamook) are priced at a premium (I'll address Intel's "inventory" situation in another post) - well above $300. Intel still has no competition here and AMD/Cyrix are only making "noise" thus far. Business PCs - these sales results are not reported because most of these are done through sales channels that the media does not have direct access to. Corporations and businesses buy higher priced, more fully featured PCs and have their own PC food chain - rewarding their best personnel with the newest hardware. I have friends/brokers at three Wall Street top line firms - and they recently gave ALL their agents brand new Pentium Pro or Pentium II machines. One company requires their agents to have a notebook PC as well - and guess what they are paying for these and whose chips are Inside? (Good guess!) Workstations - Moving rapidly to Intel/Windows NT. Cadence announced internally that all their new software WILL RUN on NT. LucasFilms is using 100% Intel Pentium Pro/II and Windows NT running Studio MAX 3D for their special effects. Avid Technologies has ported (or will soon complete) all their video editing software to Pentium/Windows NT. Accept the fact that PCs are used daily by many technical users who need and demand the last MegaHertz in performance - and that is INTEL (and a few DEC Alphas). Now consider this - the MediaGX and sub $1000 PC's bought in the past 12 months are going to disappoint a lot of people in the next year or so when they want more uumph to run Windows 98. If and when Windows NT supersedes Windows 9x, ALL THOSE subzeros wil be candidates for REPLACEMENT - NOT UPGRADES. And their owners will think long ans hard about buying on the cheap for price and NOT performance. Servers - this market is exploding - know any business without a WEB SITE or department server? Sun gets the cream of this but Intel/NT gets the bulk. And Servers have more than 1 Intel CPU per box - and that number will increase (CPU's per server) as networking increases. And you already mentioned MERCED - Big Silicon. This will be the Lingua Franca of high end enterprise server/applications from year 2000 on out. In the maen time, Intel is rapidly bringing their 0.18 micron into production readiness while the 0.25 micron process is ramping into full production in four fabs. Die sizes will drop, costs will drop, power will drop and Intel will crunch out more MegaHertzes. Profit margins will ultimately be restored and hopefully rise. Intel will also be exploiting their new Flash memory technology (Strataflash) and their networking products. This may not lead to sudden ASP jumps (Kumar's prediction excepted) but a gradual increase ASP's, starting in the 3'rd or 4'th quarter of this year. Paul

To: Patient Engineer who wrote (44027)	1/5/1998 9:35:00 PM
From: miraje	Respond to of 186894

Patient Engineer, Thoughtful post. You are correct about lowering ASP's but remember that cost of production is also decreasing substantially as the migration to .25um ramps up. I'm looking forward to hearing Intels comments on future margin expectations during the post earnings conference call. Your comment on home PC penetration being at the 45% level leaves quite a ways to go before reaching the saturation level of TV's and VCR's. And that's just in the USA. The rest of the world is a pretty big market. Regards, JB

To: Patient Engineer who wrote (44027)	1/5/1998 9:36:00 PM
From: Deepak Jain	Respond to of 186894

I have reached some similar conclusions through my analysis of the situation, and wrote something up over the weekend, which follows here. Intel's stock experienced only a modest appreciation during 1997. The stock's gain in 1997 were nowhere close to the gains it enjoyed in preceding years and it substantially underperformed both the DJIA and the S&P500 index. Recent trend behind Intel's earnings isn't particularly soothing to investors: over the last eight quarters, Intel's earnings per share in sequential order have been $0.51, $0.58, $0.74, $1.06, $1.10, $0.92, $0.88, and $0.91 (estimate). So have the earnings reached a plateau, or worse, in for a continued decline? Are there factors that might help Intel resume an uptrend in the earnings? What follows is a series of articles in which I attempt to put some recent developments in perspective and address these questions through a discussion of the following issues (since x86 microprocessors are a large portion of Intel's business and revenues, I here concentrate only on that market). The Sub-$1000 PC Phenomenon The Need for Processing Power The Competition I close by providing a personal view of the short-term and the long-term outlook for the company and its stock. (continued)

To: Patient Engineer who wrote (44027)	1/5/1998 9:38:00 PM
From: Deepak Jain	Respond to of 186894

The Sub-$1000 PC Phenomenon One of the major stories of 1997 was the rise of the sub-$1000 computers. Whether these computers are bringing in new buyers to the market, or simply displacing sales of higher-priced computers is still being debated. The truth perhaps lies somewhere in between those two scenarios. One thing is certain, however: the phenomenon of sub-$1K computers had - and most likely will continue to have - an influence on Intel's bottom line. The act that jump-started the sub-$1K market was the introduction of Presario 2200 by Compaq in February 97. Yes, there were sub-$1K computers available before then, but except for Packard-Bell (which has its own reputation problems), mostly from no-name manufacturers. An entry by the world's No. 1 computer maker legitimized this market segment, and set a trend which other top-tier vendors (such as IBM, HP, and Acer) found impossible to ignore. An obvious consequence of the rise of the sub-$1K PCs is that Intel stands to collect fewer dollars for the microprocessors in these computers than the higher priced ones. But the true impact of this phenomenon reaches much farther than just the sub-$1K category, and lies in the dramatic shift in consumer psychology. Since Joe Public can get an HP or Compaq Pentium 166 MHz MMX computer with a decent configuration for $699, how many people do you think will shell out two or three times that much for a 266 MHz or 300 MHz Pentium II. And even though businesses are less sensitive to prices than home buyers, they are not immune to price differentials and will not pay an unlimited premium to get the latest in performance. Thus this rise in consumers' expectations of what should be available at a specific price point implies that average prices for the top-of-the-line computers would also have to be lower than they traditionally have been, impacting processor prices not just at the low end, but throughout the entire product line. Indeed, surveys over the last year confirm the downward trend in average system prices. (Thus all those disputes about whether sub-$1K PCs are 10% of the market or 15% of the market are missing the bigger picture.) The dismissive attitude of some on this board about the sub-$1K computers as being underpowered or toys notwithstanding, the phenomenon is irreversible. As an Intel VP himself noted in a recent interview, "the genie is out of the bottle." Some would argue that the reduction in average system prices we are witnessing is also a result of substantial reduction in prices of other components (hard drives, memory, etc.) over the last year. That is certainly true, but it does not imply that processors are off-the-hook. First of all, rather than spending an extra $200 for a 266 MHz Pentium II over its 233 MHz counterpart today, some people might choose to improve performance by spending $50 on an additional 32 MB RAM instead. Furthermore, to the extent that the CPU assumes a larger fraction of the system cost, a reduction in prices of other components ironically creates problems for Intel. That is because the price differential between Intel's CPUs and comparable CPUs from competitors is more visible if CPU is a larger fraction of the overall cost. A $100 saving on a $900 computer appears more enticing than a $100 saving on a $2500 computer. Intel's response to the sub-$1K PC phenomenon, as embodied in the Basic PC initiative, is puzzling. Apparently, Intel would ship a cacheless Pentium II chip by this fall for these "segment zero" PCs. If recent history is any guide, at least 266 MHz Pentium II's should find their way into the sub-$1K PCs on their own by Fall 1998. By what appears to be a saving in manufacturing cost of about $20 (though with a disproportionate penalty on performance), it is not clear how successful Intel will be with this crippled CPU. Of course, it is possible that Intel is working on some variant of an integrated processor, a la MediaGX, an effort it may have chosen not to make public. Barring that, the cacheless Pentium II appears to be a case of too little, too late. Since even today Intel has the majority share of sub-$1K PCs without any specialized processor, perhaps Intel should let things take their own course and let today's upper end processors continue to migrate to the inexpensive computers in due course of time. (continued)

To: Patient Engineer who wrote (44027)	1/5/1998 9:40:00 PM
From: Deepak Jain	Respond to of 186894

The Need for Processing Power One of the reasons for the popularity of sub-$1000 PCs is a lack of mainstream applications that are demanding enough to convince the buyers to opt for more powerful systems. In earlier days, a marked difference in performance of mainstream applications (e.g., business productivity applications such as word processors and spreadsheets) was apparent between lower end machines and higher end machines. For many, such difference justified the extra expense of higher end machines. But even entry-level computers perform admirably today on those applications, and the improvement in performance on more expensive systems is apparent only to benchmarks. (As an aside, the diminishing returns are not just a matter of perception. The percentage increase in the clock speed between a 33 MHz 486 and a 50 MHz 486 chip is equivalent to that between a 233 MHz Pentium II and a 350 MHz Pentium II. Of course, we now appear to have three additional speed grades - 266 MHz, 300 MHz, and 333 MHz - in the middle.) Clearly, to sell the higher end chips, applications that are hungry for processing power are necessary. Some had hoped that the World Wide Web, with its rich graphics and multimedia content, would be such a "killer application." At least for the average home user connecting to the Internet via a POTS modem, the bandwidth and not the CPU is the bottleneck for such applications. In the business world, a significant number of computers are used primarily for word processing, presentation graphics, spreadsheets, etc., which as discussed earlier, do not always justify the extra processing power. While it is true that some people, such as game players and software developers, can certainly use the extra horsepower that comes with the higher-end processor, the question is what percentage of the market do they represent? Does this mean that I think we have reached a saturation point in terms of the need for processing power? Absolutely not. Would the average Joe require more powerful computers in the future than available today? Without a doubt. With applications like digital photography, speech recognition, and videoconferencing on the mass-market horizon, the need for higher-powered machines in the future is inevitable. However, it would be a fallacy to immediately extrapolate that into a rosy scenario for Intel, for the right question is not whether more powerful applications would become mainstream in the future (they would), but whether the more powerful applications would become mainstream in such a rapid fashion to once again begin to outpace the hardware offerings of the time. An example would perhaps best illustrate the above point. Next year's computers would be more powerful than today's computers at the same price point. Perhaps today's top-of-the-line processor will be found in next year's entry-level machines (if you doubt that, remember that a Pentium 200 MHz MMX was the top performer at this time last year, and with HP's $799 machine it is becoming the entry point today). If Moore's law continues to hold, the same will be true for the subsequent year, and so on. If digital photography takes, say, three years to gain mass-market acceptance, the entry-level machines of those days may be adequate for the application anyway, thus providing little upside for Intel. (By the way, the three year estimate is quite optimistic: just look at the acceptance of APS in the photography marketplace.) In short, the right question is not whether more demanding applications become mainstream, but when will they become mainstream. One reason for buying more than what you may need today is insurance against rapid obsolescence. However, that argument is getting weaker in this era of cheap PCs. You can buy a capable $799 PC today, and when it becomes obsolete a couple of years later, hopefully buy a similarly priced (but much more powerful) PC down the road, for a total life span longer than what $1600 would buy you today. (continued)

To: Patient Engineer who wrote (44027)	1/5/1998 9:42:00 PM
From: Deepak Jain	Read Replies (1) \| Respond to of 186894

The Competition For a significant period after the introduction of its Pentium chips, Intel enjoyed a virtual monopoly in the market for Pentium-class devices. With AMD fumbling badly with its K5 chips, Cyrix having heat problems with the earlier versions of its 6x86 alternative, and NexGen lacking a pin-compatible alternative for the Pentium, the road for Intel was fairly smooth. That situation, of course, has changed. As a consumer, the emergence of credible competition is welcome news. In fact, Cyrix (now part of National Semiconductor) and AMD deserve much of the credit for the dramatically more affordable PCs today. Granted that the best offerings of AMD and Cyrix today (a 233 MHz-class chip) are not at par with Intel's best offering at the moment (a 300 MHz Pentium II), the distance between these companies and Intel is less than it has ever been in the past few years. The emergence of such competition and the downturn in Intel's earnings is perhaps not a coincidence. Some Intel investors would claim that the competition has managed to get "only" a 10% market share (implying that somehow that much market share is not sufficient to impact a dominant player's pricing power) and that there have been previous occasions where the competitors have had an even greater market share. The difference this time, however, is that the competition is fairly close to Intel's top end and not just feeding off yesterday's chips. For an indication of the change in the competitive landscape (and the pricing power), recall that in the fourth quarter of 1996, with its competition in poor shape, Intel decided to skip regularly scheduled quarterly price cuts on its chips (and consequently had a blockbuster quarter). The decision purportedly was made so as not to disrupt the PC sales during the important Christmas season, and the practice was to continue in the future. The sales disruption argument notwithstanding, sharp price cuts were back in the fourth quarter of 1997, however. In fact, Intel even announced an unusual additional price cut (on 233 MHz Pentium II's) between the regularly scheduled quarterly cuts. Anyone who still does not recognize or admit the competitive pressure Intel is facing has to be in denial. Intel has been fortunate that Apple has been doing so poorly during this time, and a number of people who would have otherwise bought a Macintosh opted for a PC instead, thus boosting the demand for x86 chips. But offsetting this positive development has been the growing market share of the top-tier PC makers at the expense of smaller players. Such consolidation of market share amongst a handful of box makers shifts some of the pricing power from Intel to the system manufacturers. Indeed, several of them - including Compaq, IBM, HP, and Acer - felt bold enough to include non-Intel chips in their offerings. And these companies as well as other top-tier vendors undoubtedly used AMD and Cyrix as leverage to extract price discounts from Intel. Of course, there are other competitors on the horizon, such as Centaur, IMES, and Rise Technology. The window of opportunity for the competitors, however, may be short-lived. Intel's not-so-secret weapon lies in the proprietary architecture of Slot 1 (and its successors), which is patent protected and has the potential of locking out the competitors. Intel's dominant market share and its will will make Slot 1 the de facto standard within one to two years. And faced with a closed architecture, nearly all of Intel's competitors will face a tough future. The only exception could be Cyrix, which through the cross-licensing agreement between its parent National Semiconductor and Intel, has laid a claim to Slot 1. In fact, I often think that a wise move for Intel would be to not dispute that claim even if it has good legal grounds. Such a decision would keep nominal competition alive (while severely hurting AMD and preventing entry from new players), and at the same time keep FTC off of Intel's back, which otherwise may find Intel's practices anti-competitive. (continued)

To: Patient Engineer who wrote (44027)	1/5/1998 10:40:00 PM
From: Joey Smith	Respond to of 186894

Patient Engineer: re:Homes and businesses are reaching saturation. Wrong Assumption, imo. joey

To: Patient Engineer who wrote (44027)	1/6/1998 1:19:00 AM
From: Joe NYC	Respond to of 186894

P.E. Good analysis of Intel. I even agree with your price range: Buy if stock get below $60 Sell in high 70s or 80s That's the price of "victory" over AMD. In dollar terms, Intel probably paid several 100s of millions, if not billions to knock some 50 millions off AMDs earnigns. Joe