Programmable logic represents a fundamental shift for the industry and designers, and Nick Martin of Altium says their complexity should not be a barrier

It’s pretty clear that FPGAs represent the next ‘big thing’ in electronics design. Their stunning sales growth and increasing pervasiveness is hard to ignore. But just what is really going on here and how can you be part of it? To put it more bluntly, how do you avoid being shut out of what is obviously a paradigm shift in electronics design?

The conventional evolutionary view of FPGAs is that they are simply a cool new way of implementing digital logic. You know the story – better, faster, cheaper (did I mention “field configurable”), but essentially doing the same thing more efficiently. A comfortable and safe view – owned and operated by the new HDL gurus. It’s all about achieving the nth degree of optimisation, and mainstream design engineers need not apply!

But there is an alternative and “bigger picture” view, one that is disruptive to existing approaches to building electronics systems, and a threat to the conventional view of FPGAs. Looking at the history of electronics design, we can see that FPGAs are part of a long term trend in the development of electronic products – the move from 'hard' to 'soft' design.

Microprocessors were the last big paradigm change in system design. Originally used to implement calculators and then personal computers, their dynamic reconfigurability allowed them to replace existing ‘hard-wired’ digital systems and bring real ‘embedded intelligence’ to previously ‘dumb’ devices. FPGAs take this further by allowing you to build whole ‘soft-wired’ systems.

The problem is how to take advantage of this new paradigm if you don’t have the specialist skills required by current FPGA design methodologies.

Several firms have set out to find a solution which allows engineers to apply their existing system design skills to building ‘soft-wired’ systems on today’s low-cost, high-capacity FPGAs. Designing with VHDL and Verilog is certainly a powerful approach, but should not be allowed to become a barrier to getting systems built.

One approach is to package the functionality required to build soft systems into large-scale, soft-components (such as microprocessors) and then provide tools that allow you to assemble these using a familiar schematic based approach. The design can then be automatically implemented in a target FPGA device.

Soft-components are pre-verified, so the need for simulation is much reduced. The system also allows mixed schematic and RTL-level design, so RTL skills can be acquired over time if desired. The extensive nature and scale of the supplied soft-component library means that these skills may never actually be needed.

This approach allows you to build sophisticated processor-based systems on FPGAs and then create and debug the software and hardware in real-time using tried and true schematic-based building-block techniques, and conventional embedded software development techniques.

After all, board-level designers routinely create products using “black-box” discrete devices containing millions of transistors without ever having to worry too much about what’s inside them. So why should they have to learn to speak a whole new language just to harness a new piece of hardware like FPGAs?

Nick Martin is founder and CEO at Altium Limited

www.altium.com