Is Intel in Trouble?

Pages: 1 2

The Bigger They Are…

Most everyone has heard that Intel revenues are down so far this year. We’ve also heard that they are planning to aggressively attack chipset manufacturers who try to ‘copy’ the Slot 1 design. The reports from the last quarter of 1997 indicated that Intel had lost almost 6% of their market share, primarily to AMD, while early reports for the first quarter of 1998 indicate that computers with non-Intel comprised over 25% of total sales, up from under 10% last year. The lion’s-share of these non-Intel systems were sold by IBM and Compaq, with AMD processors inside. At the same time, industry analysts are saying that sales of computers and hardware were just as strong as last year, so it looks like AMD is the big winner right now. Things are looking very bleak for Intel…or are they?

Those who were paying attention also noticed that AMD reported revenues below expectations, as did Compaq. Almost all of the major OEMs are reporting inflated inventories and warning that they may have to sell systems below cost just to prevent losing even more money. So what’s going on here? How can AMD and Compaq be grabbing big chunks of the market, yet still be losing money? Is there something happening that the experts have not yet recognized?

The Pentium Killers

Almost exactly one year ago, the only serious competitor that Intel had was Cyrix – and even that was mostly just a minor annoyance, from Intel’s perspective. Intel owned almost 90% of the processor market, and almost 100% of the chipset market. The new 200MHz MMX processors were selling for over $700.00 in many cases, yet customers were scooping them up. The Cyrix 6x86L PR200, on the other hand, retailed for about $350.00, but was suffering from a very bad reputation, which prevented them from really taking advantage of their lower price. Considering that production costs of a Pentium MMX processor were somewhere between $70 and $80, Intel was making some huge profits.

Right around Q2 ’97, AMD unveiled their K6 processor, which grabbed a lot of media attention and managed to garner a fairly sizable and vocal set of followers. The processor was powerful and fast indeed, though not significantly faster than the Intel equivalent, yet somehow the processor gained a reputation for being much better than the Pentium MMX CPU. Due to very good timing, smart pricing and a marketplace ready for an Intel alternative, the AMD wave caught on very quickly. With a price of just over $500.00 in June, the K6-233 seemed like a very cost-effective alternative to Intel.

By this time, Intel had decided that the Socket 7 market was becoming too competitive and were already gearing up for the change to the Slot 1/Pentium II architecture. They announced the end of the i430HX chipset, which was just coming into it’s own. As a ‘replacement’ they offered the i430TX chipset, which was nothing more than a ‘souped up’ i430VX. The reason for basing their newest chipset on the VX was due solely to marketing reasons – the VX chipset comprised the largest percentage of sales, mostly because of price, rather than features.

In order to prevent their Socket 7 market from eroding too quickly, Intel announced a dramatic price drop for their Pentium MMX processors in late July. The Pentium 233MHz processor dropped by 55%, which of course put some pressure on AMD and Cyrix to make some adjustments. At the same time, Intel informed their authorized dealers that the Pentium II processors would provide better margins than Socket 7 processors. They promised that Intel would definitely not cut prices as drastically as they had done with the Pentium MMX processors.

I’ve fallen and I can’t get up!

The next scheduled price drop (October) saw the Pentium II processors drop over 20%. At the same time, the K6-233 was becoming very scarce in the retail channels because IBM was buying all they could for their just-announced Aptiva line of computers. While officially the chip shortage was blamed on huge purchases by IBM, stories of low yields at AMD factories began to emerge. The K6-166 was no longer being produced, and the Cyrix 6x86MX PR166 would soon disappear as well – primarily because the pricing was so low it didn’t make sense to manufacture them anymore. In addition, Intel dropped the PII 233 ‘out of cycle’ at the end of December. This was the first such ‘unscheduled’ price drop we had seen for either the Pentium MMX or Pentium II processors. The following month (January), Pentium II prices dropped as much as 30%. So much for maintaining margins!

While all of the attention was being focused upon the lack of Pentium II sales, and the increase in AMD sales, there was an undercurrent of concern by those in the industry. AMD was having serious problems getting their .25 micron technology to produce sufficient chips. Yield problems at their fabs producing the ‘standard’ .35 micron chips were causing supply problems at the retail end. Cyrix sales were also flagging. The memory market was also in a panic. A sudden surge of extremely low-priced memory caused prices to drop dramatically. U.S. manufacturers (Micron and TI) became alarmed and filed suit against several Asian chip manufacturers for dumping (which they recently won). Everywhere prices were falling very quickly, yet sales were not increasing as one would expect. Rather they were staying about the same, at best. Similar volume at lower prices means fewer dollars being spent.

While these processor battles were taking place, and all of the attention was being focused upon falling prices, another battle was being waged. In the retail channel, vendors were fighting over smaller and smaller pieces of the consumer pie. The typical ‘business’ cycle in the industry is that sales slow down after April, and then pick up again after August. In ’97, the sales slowed to a virtual crawl starting in March…and have never recovered! At the same time, more and more vendors began to appear on the Internet to hawk their wares and try to make up for lagging local sales. Soon the already small margins began to shrink even more.

In today’s business environment, cash flow means more than profit margins. If a business can generate sufficient volume of sales, even a 5% markup can generate a net profit. This is how some of the huge retailers play the game, and suddenly hardware vendors were trying to do the same thing. Customers began to realize that prices were dropping rapidly from week to week, and started working vendors against each other. Some vendors began to use somewhat questionable practices to gain customers, and reduce costs – many of which have been outlined in a previous editorial (called ‘Stupid Vendor Tricks’). Most manufacturers and vendors were biding their time, waiting for the ‘big’ Christmas season that everyone thought would happen.

Well, September rolled around but the anticipated crowds didn’t appear. The October price cuts were announced, yet sales barely hiccuped. Very late in November, the holiday buying season began – then ended just as quickly at the end of December. Many OEMs had rolled out their sub-$1000 systems in anticipation of generating some huge volumes of sales, yet they didn’t really materialize. One of the problems with the low-margin/cash flow theory is that if you drop your margins in half, but the sales volume remains the same you have effectively eliminated your net profit, and this is exactly what happened.

Now, vendors started to panic. Many had relatively large inventories but little volume. In order to spur sales, prices were dropped to just above cost. Customers would flock to whichever vendor had the cheapest price for a particular product, until they were sold out. The other vendors would basically be selling very little, and buying even less. Distributors were complaining that sales were exceptionally poor, manufacturers were complaining that profits were down, resellers were complaining that customers just weren’t buying. Customers were comparing prices from one vendor to the next, and accusing those with higher prices of ‘gouging’, not realizing that even the highest priced vendors were only getting 20% to 25% at most. The lowest priced vendors were making virtually no profit at all.

Pages:   1 2  Next »

Be the first to discuss this article!