Altium
Next generation electronics design solutions
There's nothing quite like a contentious question to galvanize views and options.
Considering what the future holds for board layout specialists is an important question in itself, but implying that these design engineers may need to ‘move on' is quite another thing. It implies a shrinking design domain, and there's something very disturbing about the concept of entropy in a fast moving, technology-based industry.
Note that the question nominates ‘specialists', and to be clear, it concerns the current view of the discipline and how that needs to change. So what's really in doubt here is how sustainable the future is for specialist electronics designers that work in an isolated design environment, and in particular, those that are dedicated to PCB design.
And the concept is not the stretch it may first seem when you take a broad view of the industry and where it's heading. With technology advancing at breakneck speed, the way engineers work and the tools they use are struggling to keep up with changes that show no sign of abating. It's these technology developments, and the demands of those who use electronic products, that set the agenda for how design is done and what the future holds.
For board-level design engineers, the path is largely set by the evolution of the very technology that PCBs host. This ultimately defines the physical and connectivity characteristics of the board design, and therefore the materials, techniques and design tools that are needed to attain those properties.
But, and here's the important bit, if the core purpose and format of PCBs change, so must the way they are designed. And the role of PCBs is indeed changing.
Where are PCBs heading?
Taking the narrow, traditionally-held view for the moment, there're a number of fundamental trends in PCB design and construction that are likely to continue
On the surface, it seems reasonable that products will continue to get smaller and smarter, driven by consumer demand and faster, high-density semiconductor devices that offer more capabilities. The PCBs used to interconnect those parts will therefore become more complex and smaller to accommodate all that capability in a compact, user-functional box.
Not too many surprises here, and the direct implication of those trends is the current evolution of advanced PCB techniques such as flexible substrates, micro-vias, high density interconnect systems and high density routing. With this board-centric view alone, it seems the future of PCB specialists is to further hone their skills to accommodate new board design technologies and techniques. It involves continuing down the same board-focused path.
However, it gets more realistic when you stand back and take a broader, more holistic view of these changes. This is a view that considers the design of a whole product, rather than one that just extends traditional thinking from a board design perspective. It comes from not looking at electronic product evolution through PCB-focused goggles.
The changes that have occurred to make smaller products possible are mainly driven by the evolution of semiconductor technology, while PCB technology and techniques have adapted to capitalize on this. Large-scale devices in high-density packaging continue to offer more functionality and require less supporting chips, which reduces the component count and number of board-level interconnects. So in this sense, PCB designs are actually becoming simpler.
The revolution delivered by programmable devices such as FPGAs has taken this change to the next level by introducing 'soft' hardware design and a programmable SoC approach. Along with the low non-recurring engineering (NRE) costs, design flexibility and ecosystem connectivity advantages this offers, board complexity is further reduced as large sections of physical hardware are moved into the soft domain of programmable logic.
Couple the adoption of programmable devices (and ‘soft' hardware) with the growing emphasis on software to define design IP, and a wholesale swing towards soft design becomes apparent. The result is a shift away from hardware as the dominant factor in product design. Board level design is becoming simpler and more modular, while the physical configuration, connectivity and construction are now dominated by other design factors.
An example of the influence of other factors is FPGA connectivity. One of the unique aspects of FPGAs is that the device pin configuration itself is programmable. Traditionally, this is the bane of board layout engineers who not only have to fan out hundreds of densely packed device pins, but also deal with hundreds of pin assignments that are determined (and changed) in the separate FPGA design domain.
While from a purely PCB design perspective this creates design complexity, a more holistic view is that variable device pinouts can be used to advantage when the PCB and FPGA design domains work together. More specifically, a PCB designer might rearrange the FPGA pin configuration to simplify routing, and the changes are directly reflected in the FPGA design space where the place and route tools automatically reconfigure the FPGA to match.
Again, the board design is simplified, and in this case another domain is involved. This process is viable in a fully connected PCB-FPGA design environment, where it can even be taken to the next level by moving problematic routing paths inside the programmable fabric of the FPGA. Here, the FPGA place and route tools solve board routing challenges, but only if PCB-FPGA development exists in a common design environment using a single, shared model of the design data.
While those concepts are very compelling, the revolutionary opportunities provided by programmable logic are set to expand even further as patents on the technology expire. Consider the design possibilities when a hardware device, let's say a processor, incorporates a level of programmable logic in its structure.
In this case, how that processor connects and interfaces with the other parts of the design are now programmable. The device can be programmed to include the support devices and peripherals needed in your particular application, and as with traditional FPGAs, the pin configuration can be optimized for the board layout. The result is fewer parts on the board, and simpler interconnection paths.
It's not too much of a stretch to take this concept to a point where every device in a design could be configured to directly interface with the others. Only simple, direct electrical interconnections would be needed, and the devices effectively plug together - perhaps even literally.
When taken to this logical extreme, board design devolves into arranging the physical support for electronic and mechanical parts. It would contain comparatively few electrical paths, and its properties and shape would be entirely determined by mechanical considerations. The ‘circuit' boards of the future may well be designed in the MCAD space.
Other future developments are bound to reinforce the move way from the traditional view of a PCB containing mass connectivity. Electronic components such as LCD screens and buttons embedded in a product's case, configurable conductive plastics, and even dedicated programmable devices that take charge of all routing issues are likely to reduce the emphasis on ‘pure' board design.
Embracing change
You can see where all this is heading. Circuit boards in future products will take on a variety of unfamiliar forms that are likely to move away from mass connectivity, but their development is certain to be fundamentally influenced, or even dominated, by the other domains.
The swing to programmable devices simplifies board design, the intimate role of the product's mechanical design dictates its form, and software development defines its electrical configuration. Physical electronic hardware is becoming simpler, and no longer determines the unique differentiating elements of a product design. Conversely, the essence of today and tomorrow's designs (the unique IP) is defined by application software and the soft hardware it runs on.
In the overall sense, isolated highly-specialized electronics designers are at risk as the design disciplines interact and converge. Note that the emphasis here is on ‘isolated'. Specialist skills are valuable, but can no longer exist as disconnected silos of electronics development. Electronics design is just not going down that path.
So is PCB design itself dead? Certainly not, but it will radically change as programmable parts flourish and external influences continue to dominate board design. What's not assured is the future of those who specialize in PCB design as an isolated process and hang on to a sectarian view of the discipline.
As the traditional divisions between the design domains blur and the nature of product development changes, board-level design needs to be viewed as part of a single holistic process - and not an entity to itself. Effective and transparent design collaboration between disciplines is currently important, but will quickly become essential as electronics design is approached at a holistic level rather than piecemeal. And that means engineers will need the skills and ability to reach into and influence unfamiliar parts of the design process, starting with considering what the final user experience is supposed to be.
Paths for the future
So as a board level engineer, does that mean you need to plod through extensive training in an arcane hardware description language (HDL) to work with the programmable hardware in your designs? Well, no. Re-skilling to embrace the other design domains might seem daunting and impractical, but only if you view those domains in the traditional way - as highly-specialized and separate processes.
The answer lies in adopting design solutions that abstract the design processes to a higher level while automatically dealing with the underlying complexity. When these systems are part of a single design environment that uses a single pool design data, working in other design domains becomes practical.
Design engineers can then build on their existing skills and work as system designers rather than isolated specialists. With FPGA design for example, high level design processes that use schematic entry or graphical signal flow allow engineers to use their existing hardware skills to develop or modify embedded soft hardware. This is the next step beyond collaborative processes such as FPGA pin reallocation across the PCB-FPGA domains.
High-level design systems also offer the promise of expanding the collaboration and influence of application software developers. If the design system provides sets of software layers and drivers that handle the low-level complexity of hardware embedded in an FPGA, software developers can use general hardware design skills to create whole SoC systems to run their applications.
The theme here is that when advanced, high level design processes exist in a single product development environment, all engineers can expand their existing skills to easily collaborate with and work within traditionally unfamiliar design domains. The door is then open to harness and develop their existing engineering talent by moving towards a holistic system-based approach to electronics design.
For this to be possible, it's critical that the design environment encompasses the design domains at a fundamental level and uses a single model of the complete design data. Note that this is profoundly different from a collection of ‘integrated' design applications that simply interconnect by passing data between each other.
With that traditional approach, higher levels of design abstraction might make design more accessible in a particular domain, but will only increase the overall design complexity when the specialized data are passed to the other domains. High level design systems are only practical when they can permeate though all the entire design process by working with a single, unified pool of design data.
In short, the future can be bright for all hardware-level engineers, including those that currently specialize in PCB design. But, and this is a very important but, only if the changes in the electronics product design landscape are recognized and embraced. These changes show a fundamental shift towards soft-centric design, reduced emphasis on physical (as opposed to programmable) hardware, vastly increased interaction between the so-called traditional design domains and a focus on creating ‘connected'' products.
By adopting systems that raise the abstraction of the design process within a single design environment, engineers can extend their current set of skills to collaborate with other engineers and even play in their domain. The result is an electronic product development design environment that makes the most of engineering talent and promotes design exploration.
Specialist designers are then free to break out of the traditional isolation barriers and creatively contribute to the design in a way that considers the product's competitive value, and the end user experience. This holistic approach to product design allows them to reclaim their competitive advantage and place in the future, and is also the path for all designers in creating the unique, connected product designs of the future.
