A few years ago, I watched the movie Groundhog Day. In it, the lead character played by Bill Murray was stuck in a time warp and subjected to repeating one day of his life over and over until he learned some basic lessons. Over the past years, we've been living our own High-Tech Groundhog Day. But what are the lessons we must learn?
The first lesson: Don't get carried away in the good times and don't give up on the future in bad times; the market will come back.
The second lesson: Just as lightning never strikes in the same place twice, the next boom never comes in the same place as the last. Where will the next boom be? Like most, I don't know for sure, but there is one thing that I am sure of. During the downturn, the most important thing we must do is to place our bets on the next upturn.
The last and most obvious lesson: Don't run out of cash. The best investments and most important decisions are made during downturns. But you have to survive the downturn to make those decisions count.
So, what's the right bet for the next upturn? For two decades, the holy grail of the PLD business has been speed, cost and density-build a faster, cheaper, bigger FPGA and the world would beat a path to your door. Today, we find that, for most common, commercial applications, most PLD manufacturers offer products that are big enough and fast enough to do the job. As a result, we've seen the emergence of a new set of criteria for choosing a PLD. While price remains a significant criterion, four new requirements have emerged: power, security, reliability and footprint. Increasingly, customers are demanding cost-effective FPGAs that offer greater security, lower power consumption, reduced board space and protection against firm errors.
To that end, we are seeing the resurgence of the $5 FPGA. For most of the last 30 years, semiconductor manufacturers have made a business selling products in the $5 to $10 region. Aided by the telecom hysteria of the late '90s, PLD vendors were exempted from this rule and were able to cash in selling many products with prices in the several-hundred-dollar region. Those days seem to be on the decline. This economy is forcing designers to look for solutions with ever-lower prices. At the same time, increasing ASIC tooling costs and time-to-market pressures force the choice of a programmable solution. The result? The emergence of the $5 FPGA as a mainstream market.
As an industry, we need to learn from history or we'll be stuck reliving these dog days. Here are the lessons as I see them: don't give up on the future, don't run out of cash and lightning never strikes in the same place twice. The next bolt? In my view, the $5 FPGA.