Intel CEO Craig Barrett this week collectively took his employees to the woodshed, chastising them in a companywide memo for a spate of recent chip production glitches and exhorting them to move quickly to fix the problems.
"It is part of our culture to address our problems with honesty and to resolve to fix them," Barrett writes in the memo, excerpts of which were provided to us by Intel. "Our business is complex, and we have set high expectations for ourselves. Therefore, it is critical that everyonefocus intensely on actions and attitudes that will continue Intel's strong track record of technology leadership."
Just what problems is Barrett referring to? In the past six months alone, they include the discovery of defects in Grantsdale (a chipset for use with the Pentium 4 processor), the scrapping of the planned Tejas processor because it dissipated too much power and ran too hot, and delays to the Dothan Pentium M mobile processor.
What's the root cause of such a sour streak?
Intel officials tell me it's because the company has been too aggressive in making commitments to its customers to ship chips as quickly as possible, and to push its technology roadmap forward so that OEMs and resellers can stoke sales demand with ever more advanced systems. Having bitten off more than it can chew on too many fronts, Intel has then fallen short.
However, they say the problems don't have anything to do with Intel's imminent migration to its advanced 90-nm semiconductor fabrication technology. That's good news, because the move to 90 nm is a critical transition that's an important component in Intel's plan to shift its product lineup to multicore processors.
Multicore involves the placement of two separate microprocessors on a single chip. It cuts power consumption, as compared to single-core processors of equivalent performance. Intel plans to roll out multicore chips in 2005; by 2006, it expects all server CPUs and more than half of all the notebook and desktop processors it ships will be multicore. Any glitches in that transition could seriously affect public perceptions of Intel.
For now, Barrett is making the right move by taking the high ground getting out in front of any potential problems.
As an experienced executive, he's doing it via very careful language, attempting to position the problems as part of a good news, bad news scenario.
"There are many reasons for these [product delays and manufacturing issues], but in the end the reasons don't matter because the result is less-satisfied customers and a less-successful Intel," Barrett writes of the difficulties. "I believe, as you do, that this is not the Intel we all know and that it is not acceptable."
Then, reaching for some positive spin, he stresses Intel's solid financial performance.
"This past quarter we did achieve $1.8 billion in profits [up nearly 100 percent from 2003] as well as higher revenue growth than you'd expect for this time of year, and we gave an outlook for $8.9 billion in revenue for the next quarter -- an all-time record if we reach it," the memo says. "But this just makes our recent problems all the more disappointing because of what we could achieve if Intel were performing well in all major aspects."
Most important, Barrett appears serious about putting an end to the problems before they get out of hand.
"We are starting to put in place the indicators, reviews and management attention to start to turn these problems around by ensuring good planning, staffing and program management," he writes. "This will not be a short-lived focus; we have plans to continue to review expectations and performance in the future."
Clearly, Intel can't afford to take its eye off the ball. Thursday evening, in the wake of the memo, Intel spokespeople called us to say that the 4-GHz version of the Pentium 4 processor, which was originally set to ship in the fourth quarter of this year, would be pushed back to the first quarter of 2005. Intel needs more time to ramp up to volume production.
Taking a page from the rationale for the memo, the announcement of delay now appears to be a pre-emptive move to ensure that Intel isn't stuck later on with a commitment to a deadline it won't be able to meet.
In the technology business, fast is the only speed we know. But now, Intel is saying that slower and steadier will have to be the prescription for at least a little while.
As Barrett sums up in his memo: "Intel has been in tough spots, and we have always gotten through them because everybody focused intensely on the challenge -- and we have always emerged a stronger company as a result."