| In the industrial market, designers have a significant incentive to get their products to market quickly to maximize revenue and time-in-market. Every week that a product is not being sold represents lost revenue, increases the product's market risk and lowers the chance of success. The difference between getting a product out and selling today vs. a three- to six-month delay can be the difference between profitability and product failure. |
For example, if a typical sale brings in $1,500 and a company expects to ramp sales to 100 per week, a three-month design delay could cost more than $1 million in lost revenue. That level of lost revenue can bring a lot of unpleasant attention to a project. To overcome this time-to-market challenge, industrial designers are looking for fast, flexible, reliable and easy-to-design solutions for their products.
Not surprisingly, forecasts indicate that the programmable logic content in industrial designs will double to more than $1 billion during the next four years. Coupled with standard microcontroller instruction sets implemented as soft intellectual property (IP), field-programmable gate arrays offer a flexible and fast time-to-market platform for industrial designers. This was apparent in a recent industrial wireless communication design where an application-specific standard product (ASSP) and then an ASIC were initially considered. When time-to-market, implementation flexibility and future obsolescence were considered however, the design team chose to implement the project with an FPGA.
In addition to time-to-market pressures, there are a number of other reasons why designers are switching to programmable logic devices for value-added functionality in industrial designs. Modern process geometries are enabling a new class of programmable logic that delivers more logic at higher speeds with faster I/O and lower price points, allowing FPGAs to be used in embedded applications that were once addressed by ASICs or ASSPs. No longer limited to simply implementing system glue logic, today's high-function FPGAs are being used as system-on-chip platforms.
Using FPGAs, designers can develop a custom solution without the nonrecurring engineering (NRE) charges or the fabrication and assembly time delays typically associated with ASICs. Designers do not have to deal with the unwanted functionality provided by ASSPs or the inevitable silicon spins associated with ASICs.
FPGAs offer a low-risk, quick time-to-market solution that industrial designers can easily modify when they need to make changes, fix bugs or create product derivatives at some point in the future.
Another reason for the increasing use of FPGAs is the dramatic increase in the breadth and number of available IP blocks that can be easily programmed into the device. This includes an assortment of standard functions that are widely used in industrial applications. These preverified and thoroughly tested IP blocks have been optimized for use in programmable logic and enable designers to quickly assemble a system and program it into an FPGA.
The IP cores are typically available as a netlist or RTL source so designers can use a core quickly without change, or easily configure it to their design requirements. Also, the cores' compatability with the popular 8051 instruction set allows designers to leverage their experience with the microcontroller architecture and take advantage of the huge volume of code and tools that exist for it, reducing development time. And often, the cores have additional features, including on-chip debug capabilities, that ease system debugging when the core is deeply embedded, thereby helping designers get their products to market faster.
The benefit of using FPGAs with IP was evident in the recent design of a modular wireless industrial network for use in high-noise factory environments and building automation. Initially, the design team looked at using discrete ASSPs, but quickly determined that it would not be possible to use them and obtain the right mix of desired functionality or meet the necessary size and power requirements. In addition, during the project cost analysis, the team found that at the volume projected for the module, the ASIC and FPGA device costs were similar but that the FPGA did not require an NRE. As a result, the team decided to use an FPGA solution. Additionally, because the FPGA vendor had most of the IP needed for the project, the design team only had to develop a small amount of differentiating IP. At this point, the advantages of FPGAs became evident.
Using predeveloped and -verified IP in an FPGA reduced the design cycle by as much as six months, allowing the team to beat its schedule and deliver the product to market sooner. In addition, because of the flexibility of the FPGA, team members found that they could customize the modules based on the application and specific needs of their larger customers. Using a reprogrammable FPGA also enabled the company to upgrade products in the field, without replacing the board, by simply reprogramming the device. This ability reduced the cost of ownership, increasing the value for the company's customers and market demand for the product.
FPGAs offer significant advantages over ASSP and ASIC solutions in terms of time-to-market, implementation flexibility and future obsolescence concerns. In addition, because many industrial applications never achieve high volumes, FPGAs provide a cost savings over traditional ASIC implementations. They appeal to industrial engineers because they give designers the ability to quickly program functionality and test the product in the application, and then reprogram the functional specification changes.
FPGAs are the modern solution for industrial designers that enables them to take advantage of the standards with which they are familiar. FPGAs allow these designers to get their products to market quickly, maximizing time-in-market and revenue.
Mike Thompson (firstname.lastname@example.org) is marketing manager for intellectual property for Actel Corp. (Mountain View, Calif.).