In yet another devastating body blow to the company, Intel (Nasdaq: INTC ) announced it would be canceling its 4-GHz Pentium chip. The semiconductor bellwether said it was switching gears, changing its focus from increasing microprocessor clock speed to a more ephemeral measurement of performance in multitasking, security, and multimedia.
Archrival Advanced Micro Devices (NYSE: AMD ) had previously shifted its focus away from ever greater clock speeds in 2001, and Intel's decision ends 25 years of clock speed improvement that has seen the speed of its microprocessors increase by 750 times. It's a move Intel says it has been preparing for all along.
It might be more believable had not the company been touting the fact that it would be soon releasing a 4-GHz chip, or if it had not had similar miscues and changes of direction all year long. Fool contributor Tim Beyers has made a career this year of tracking Intel's missteps, mistakes, and flubs. Let's see:
- At a time when AMD is hitting on all cylinders and gaining market share, Intel was announcing disappointing third-quarter numbers.
- It delayed the introduction of a laptop computer chip.
- A manufacturing flaw in one of its newest chipsets could cause computers to crash.
I could go on, but it becomes too painful after a point. The current announcement seems to recognize that it has finally reached a hurdle too high. When the Pentium 4 was released in February, Intel said it would release a 4-GHz version by year's end. That was then pushed back to March 2005 when it said it had concerns about meeting customer demand. Canceling the 4-GHz goal altogether is a realization that, as IBM (NYSE: IBM ) has noted, sacrificing power efficiency for the sake of speed no longer makes sense. Undoubtedly, this puts an end to rumors of the 65-nanometer, 10-GHz "Nehalem" Pentium 8 chip.
To replace the chip, Intel will instead offer a 3.8-GHz chip with a twin billing of memory known as "cache." Cache stores frequently used data in a location that helps the computer access it faster, rather than having to search the main memory banks. But this is not without its own set of problems. The larger, dual-core chip may be more expensive to make, and it may not be able to sell them at an equivalent price.
It would, however, help Intel use its expanded capacity. Having spent some $20 billion on 90-nanometer and 300-millimeter chip designs over the last four years, it is faced with a weaker-than-expected PC market. The company has warned investors it will take an underutilization charge when it reports fourth-quarter earnings later this year. Yet the larger size requires Intel to produce more silicon wafers, thus spreading out the costs over a larger number of chips. It could help stabilize margins.
Without question, Intel is a giant struggling in a changing landscape. IBM and Sun Microsystems (Nasdaq: SUNW ) already make dual-core chips; AMD has proven a formidable adversary; and even Texas Instruments (NYSE: TXN ) holds sway in the mobile phone market.
Fool contributor Rich Duprey holds sway with Coors Light and Krispy Kreme doughnuts. He does not own any of the stocks mentioned in this article.