The emerging Configurable System-on-Chip (CSoC) market is quickly establishing itself as one of the most exciting sectors in the semiconductor industry. Fueled most notably by communications applications, the CSoC sector is poised to explode into a $50 billion market by the end of the decade. Dataquest's estimates tell us that CSoCs will account for about 80% of all SoC-class semiconductor devices shipped within ten years.

As this market begins to mature, designers will need to start choosing strategic vendor partners, in part by deciding which will succeed and which will fail. The winning vendors will be those with product focus and a comprehensive platform strategy for hardware and software. CSoC technology is actually already available. Current CSoC devices combine a full fixed-logic system (at the least, a microprocessor, bus, and memory) with embedded programmable logic. The devices deliver the time-to-market benefits of a solution-oriented ASSP with the flexibility of a programmable-logic device (PLD).

In 1998, Triscend pioneered the CSoC market with its E5 CSoC family, based on the 8-bit 8051 MCU. The company is now shipping its second CSoC family, the A7, based on the ARM7TDMI RISC processor. Other vendors have since announced CSoC devices. Altera has announced the Excalibur family of CSoC devices. Along with Nios, an internally developed embedded processor, Excalibur will embrace an ARM9 processor in one family and a MIPS processor in another. Altera expects to deliver the first Excalibur silicon early in 2001.

Xilinx very recently announced its Platform FPGA initiative, which will embed hard cores into an FPGA. The first Platform FGPA devices will put an IBM PowerPC 405 microprocessor hard core into a Virtex II. Another programmable-logic convert is QuickLogic, which has announced its intention to produce a CSoC device that incorporates a MIPS processor. QuickLogic, which calls its configurable devices Embedded Standard Products (ESPs), embeds a hard IP core with a programmable logic array, coupling the core and array with a high-speed interface.

Initially, selecting CSoC devices will be straightforward for designers. As these devices begin to hit the market, they will be sufficiently differentiated from one another by the embedded microprocessor, on-chip bus, amount of on-chip programmable logic, and development tools. Most designers will be able to easily determine which CSoC part is the right fit for their particular application. The various CSoC devices differ from one another on many fronts, including microprocessor selection, price points, development software support for drag-and-drop configuration of the programmable logic, amount of programmable logic, and internal busing scheme.

Most CSoC applications will probably find a natural fit to one of these families. Over time, however, the number of products and the complexity related to these products will grow. In addition, CSoC devices will increasingly be used as critical components in end equipment. As this trend develops, many companies will find the need to begin forging strategic relationships with the leading vendors. The question is, "who will lead the market?"

As in most high-technology sectors, the winners in the CSoC arena will likely be the players that:

1. Establish an early lead
2. Are highly focused
3. Have a key competitive advantage.

The influence of the first two factors in gaining market leadership is well established in many high technology markets. The early leaders usually prevail, in large part because they are completely focused in this one area and, despite later pressure from bigger players, can outperform their distracted competitors. In addition, competitive advantage will be a crucial deciding factor in determining the winners of the forthcoming CSoC battles. I argue that the winners will be the ones that have the platform infrastructure to respond most quickly to the needs of customers.

The fundamental value proposition of CSoC devices is time to market, a critical issue directly related to the degree that the application requirements are delivered out-of-the-box. This means that CSoC vendors will need to put an emphasis on understanding application spaces and quickly delivering system solutions that deliver faster time to market. This is very similar to the challenge faced by ASSP vendors.

On the other hand, the high-end PLD market has historically put its emphasis on product operations to keep PLD architectures migrating to small process geometries as quickly as possible to reduce prices and improve performance. But for CSoC devices, the embedded programmable logic is just one part of the system, and the winning CSoC solutions will be those that successfully integrate all system elements to deliver a solution that balances features, flexibility, and cost.

Thus, the CSoC vendor that can most quickly translate market opportunities into well-rounded product solutions will have the best chance to prevail. Perhaps the best way to introduce CSoC products quickly is to create and leverage a baseline CSoC core platform that contains a system-oriented, embedded programmable-logic and backplane bus. This core platform can serve as a starting point to quickly create varying CSoC devices that may vary in the microprocessor and fixed-logic functions. Legacy PLD architectures were built for use as standalone devices, and thus will not be well positioned to act as such a core platform. This is the biggest hurdle that established FPGA vendors moving into the CSoC space will need to overcome.

In predicting which vendors will succeed in the CSoC era, look for companies that have developed a core platform approach to their CSoC architecture. A platform approach likely means that they can effectively address the logistics issues of rolling out CSoC devices with a template that allows them to quickly meet market needs with silicon and software. Such a template should include a dedicated bus, advanced debugging capabilities, and a way to easily retarget IP across CSoC devices, all to enable the shortest possible product-development time.

We are at the dawn of an exciting new era for semiconductors—the Configurable System-on-Chip. As the industry grows into its potential, vendors will present designers with an increasing variety of CSoC products. These designers should pay careful attention to which vendors have established the early lead, are focused, and have a platform approach as their competitive advantage for quickly rolling out true solutions. These vendors will make the best partners for the long haul.

Chris Balough is the senior director of strategic planning at Triscend. He is responsible for defining and developing the company's product roadmap. Prior to co-founding Triscend, Mr. Balough served in various marketing and sales roles at Xilinx and Texas Instruments. Mr. Balough holds a BSEE, summa cum laude, from Ohio State University, and an MBA from the J.L. Kellogg School of Management at Northwestern University.