FPGA Gurus

President Barack Obama, in his Feb. 24 address to Congress, described what hopefully happens after credit markets thaw. Unfortunately, his message must be tempered by what is happening today. Bolaji Ojo gave us an apt and unsettling description in a Feb. 24 EE Times dispatch of how semiconductor and systems companies alike are being poisoned in the toxic-assets crash. And the icing on the cake on Wednesday was a set of back to back forecasts from IDC and Gartner, predicting that semiconductor revenues would experience a 22 or 24 percent drop, respectively in 2009 relative to 2008.

While we often hear that a global market without credit is like an engine running without oil, the analogy may not be apt because the application of new commercial credit can make an economy run again. When pistons seize up, a car’s engine is ruined. When credit seizes up, an economy may be severely broken when new commercial paper is available, but it is not irretrievably ruined.

So what can those companies involved in an electronic supply chain do when credit availability seizes up? We can look to the loose talk about local barter economies for clues. The first realization is that those companies with some cash reserves are best positioned to continue to supply customers, whether on the IC, board, or OEM system level. If you are a small company that has to borrow to make payroll, you may be temporarily out of the game.

In a temporary era of reduced credit, you may still have the opportunity to participate in a global supply chain, but the notion of “just-in-time manufacturing” flies out the window. In many cases, products like FPGAs and ASSPs may have to move through distributors or through well-established customer chains, and often those opportunities may be strictly domestic. In short, semiconductor supply is reduced to a short-term billing and cash-on-the-barrelhead industry, and this may mean that IC customers can only afford to work on next-generation device prototyping, because moving a product to production is a daunting task.

There will be exceptions. Will Tibco's unique FPGA-based hardware for messaging be a success as the banking industry recovers, for example? Will new 700-MHz handsets have any opportunity to catch fire in 2009? If so, we will have to go back to those ancient 1980s notions of semiconductor allocation and oversupply. The 45-nm giant CMOS fabs of Asia may have to go through boom-and-bust production that many thought was gone in the semiconductor world.

The only alternative to such seat-of-the-pants supply chain tinkering will be to place a company in hibernation for one or two years. But there is a danger in suspended animation, beyond the concept of being out of sight and out of mind. If a company becomes too flat in management during 2009 by laying off too many people, decisions will be made about its marketing and branding that are only signed off by one or two people, operating without proper checks and balances. Perhaps it’s better to remain partially alive, using a 21^st-century version of a barter economy, than to fall into such a deep slumber that something happens under the corporate name that all its employees will regret when the pistons start turning again.