Chip vendors are finally admitting that this downturn will not end soon. In the network-processor (NPU) market, AMCC and Vitesse have undertaken sizable layoffs.
The impact of those layoffs varies. We expect little slowdown in AMCC's NPU and traffic manager (TM) plans; Vitesse, in contrast, has essentially given up on those two markets.
With $1 billion in the bank, AMCC can afford to be aggressive in its spending. Yet, the company chose to lay off 25 percent of its staff to avoid tapping that bank account. Many of these reductions, however, came from closing an old bipolar fab and consolidating its two switch-fabric teams (a legacy of the MMC and Yuni acquisitions) into one.
As a result, across-the-board cuts in AMCC's core NPU and TM development teams were limited. Some future products are likely to slip a quarter or two, but the impact on AMCC's position will be modest. We expect the company to continue to lead those markets.
Through its own acquisitions, Vitesse had assembled a product portfolio that rivaled AMCC's, including physical-layer devices, NPUs, TMs and switch fabrics. But with Ethernet NPUs and ATM traffic managers that were not compatible with each other or with the company's fabrics, Vitesse's portfolio lacked the coherency of AMCC's.
Vitesse also lacked AMCC's cash position, placing it in more dire straits. While AMCC was mainly shedding fat, Vitesse was forced to cut to the bone. Vitesse canceled development of all future NPUs and traffic managers, laying off or reassigning the development teams that came in through the Sitera, Xaqti and Orologic acquisitions.
Ironically, Vitesse's moves came just after the successful sampling of its IQ2200 NPU, the first device compatible with both its PHY and fabric products. Although the company will continue to sell the IQ2200, I doubt that many customers will be interested in an NPU with no growth path and limited support.
Vitesse plans to focus on its more successful PHY and fabric products-the company's historical strengths. Yet, the lack of a strong NPU leaves a hole in the middle of the company's offerings.
AMCC, in contrast, remains well-positioned across the entire range of line-card silicon. By shedding an inefficient fab and redundant personnel, AMCC can conserve cash and prepare for profitable growth once the recovery begins.
Linley Gwennap (www.linleygroup.com/npu) is Founder and Principal Analyst of the Linley Group.