Don't know if this was posted yet.
September 25, 2001
TO MY FELLOW EMPLOYEES:
The worldwide semiconductor industry enjoyed above-average growth in 1999 and spectacular growth in 2000. In retrospect, it is now clear that the levels of Information Technology expenditures, which drive semiconductor demand, were a bubble and are unsustainable.
The semiconductor industry expanded capacity and ramped production in anticipation of continuing robust demand that was not forthcoming. The ensuing inventory glut, the slowing economy, and severely curtailed capital spending - particularly in the communications and networking sectors - are shrinking demand for semiconductors at an unprecedented rate. Indeed, the latest estimates are that semiconductor demand will decline by up to 35 percent year-on-year as the tragic events of September 11 and the impact on consumer confidence dispel hope for any meaningful upturn in the fourth quarter, historically the seasonally strongest.
This is the sharpest downturn in history for the semiconductor industry. For as far back as the industry has been keeping records, AMD has outpaced the industry in sales growth. This year, 2001, will be no exception, but in this case, it will be by shrinking less than the industry does!
At the sales levels we currently anticipate, we will not be profitable as presently structured. We can't wait for a general economic recovery or a strengthening of semiconductor demand sufficient to return us to profitability. Our strong balance sheet will enable us to endure the difficult times, but only if we take decisive actions to improve our financial performance and to focus our organization around our two most promising opportunities - flash memory devices and PC processors.
Accordingly, we are taking steps to give our Memory and Processor business units direct responsibility for both front- and back-end manufacturing and technology. Our technology development efforts will be more focused. We will communicate specific plans regarding these organizational changes and additional measures aimed at improving our business focus and attuning our cost structure to the current business realities. Additionally, a significant slowdown in demand for products from our foundry services clients has led us to a decision to close our two oldest manufacturing facilities in Austin, Fabs 14 and 15, in the second quarter of next year. Dramatically slower growth in unit demand than previously anticipated also requires that we reduce employment in our overseas back-end operations. We will also reduce our workforce in other manufacturing operations as well as administrative support associated with these facilities. These actions will result in a reduction in our worldwide work force of approximately 2,300 positions, or approximately 15 percent. Approximately 1,000 of these positions are associated with closing the fabs in Austin, and the remainder will result from realigning and restructuring back-end activities in Penang, Malaysia.
Consistent with our values and culture, we will treat employees affected by these actions with fairness and respect.
The semiconductor industry is cyclical. The "down" part of the cycle is always characterized by sharply contracting prices. We must aggressively reduce costs for both our flash memory products and our PC processors in the harsh pricing environment we anticipate through next year.
A number of initiatives that we are pursuing in our joint venture with Fujitsu and the advent of our innovative MirrorBit technology should enable our Memory Group to extend its industry-leading position and post profits next year even in the face of very aggressive pricing.
Our award-winning AMD Athlon processor family, including our AMD Duron processors, has established AMD as the only real alternative to the Intel monopoly. We have gained more than 10 percentage points of market share to more than 22 percent unit share of the world market. The design excellence and architectural superiority of our seventh-generation processors compared to Intel's Pentium 4 are especially evident in the fact that our die size is only 59 percent of Intel's in the same process technology. This cost advantage is maintained as we move into succeeding technology generations. Competition from AMD has taken 15 percentage points out of Intel's operating profit margin over the last year.
Intel's long-term response is to invest $7.5 billion to move to 300-mm wafers and migrate to 130-nanometer technology before we can to offset our cost advantage. This response will fail. By the end of next year, 100 percent of our production in Fab 30 will be on 130-nanometer technology. Their near-term tactics are to subsidize their higher-platform-cost Rambus memory solution, to discount P4 in the consumer market with an inferior SDRAM solution (described as "a Porsche with bicycle wheels"), and to lavish market development funds on (i.e., "to bribe") and to intimidate beleaguered customers to stay with Intel. They are heavily promoting raw clock speed rather than delivered performance in an effort to drive us from the marketplace. This response will also fail. AMD will endure. Truth is on our side. As Abraham Lincoln said, "You can fool all of the people some of the time, and some of the people all of the time, but you cannot fool all of the people all of the time."
Intel is terrified at the future prospect of competing with our eighth-generation Hammer family of processors. We believe Intel recognizes that they have made an enormous blunder with their approach to 64-bit computing. Intel's Itanium solution will require users to rework or replace every piece of software they have. The AMD x86-64 solution will deliver unsurpassed performance while enabling customers to run legacy software and migrate to 64-bit computing at their own pace. We expect that the marketplace will embrace our approach. We will successfully overcome the competitive obstacles. We have the people, products, partnerships, plans, and resources to execute our own winning strategy, which is based on outsmarting rather than "out-muscling" the competition.
Hammer is the future! We can't let the bad guys win!
The industry needs and wants competition. Our smart technology embodied in Hammer, with its attendant smaller die size and intrinsically lower costs, will change the competitive landscape. The small die sizes of our current seventh- and forthcoming eighth-generation processors will enable Fab 30 alone to produce more than 50 million processors a year!
Next month we launch the enhanced-performance, seventh-generation "Palomino" core for desktops. Palomino-based platforms will deliver extreme performance unsurpassed by anything on the planet. Early next year we launch the 130-nanometer "Thoroughbred" core, opening additional new opportunities to us. By the end of next year we launch the eighth-generation, 64-bit "Hammer" for the desktop. We are going to move quickly and decisively to improve our competitive capabilities and return CPG to profitability.
We will not impair R&D efforts or new product development, the lifeblood of the enterprise.
The spirit of AMD will endure.
A virtual gorilla can respond to changing circumstances faster than an 800-pound gorilla.
The adventure continues.
God bless,
Jerry Sanders |
