Why Hardware Still Matters
By Richard Goering
Given the increasing industry focus on embedded software development, and the ongoing discussion about application-driven system development, one might ask whether hardware still matters. This topic came up at the recent ARM Technology Conference, and the short answer was “yes.”
A panel discussion posed this question: “If the whole point of systems development is the user’s application experience, are we evolving to the point where there is no reason to spend money on differentiating the underlying hardware, and where we’ll produce some small, generic hardware platforms and focus all our efforts on software and firmware?”
The best answer, in my view, came from the self-described “software guy” on the panel—David Rusling, CTO of Linaro. If we end up with a single generic hardware platform, he said, the result will be “software with no trace of innovation. The embedded and mobile space is one of incredible diversity and innovation, and from a software perspective you need to exploit all that diversity.”
I heard a related comment at the recent Silicon Integration Initiative (Si2) OpenAccess conference. Charlie Huang, Cadence chief strategy officer, noted that “the system and the user interface has become the focal point.” However, he added that “silicon is still the substrate. Without the steady progression of process nodes, we’d still be running Windows on 640k.”
The era of application-driven system development is clearly upon us. Pursuing a new business model that promises a continual revenue stream, electronics OEMs are now differentiating on the basis of “apps” rather than cool features in silicon. But those very same “apps” place a lot of new demands on silicon. Aggressive requirements for bandwidth, storage, battery life, low power, and anywhere-anytime connectivity cannot be met with “small, generic hardware platforms.”
We are heading for one of two scenarios. In the first, software applications take full advantage of advanced features in fully optimized, high-performance, low-power hardware. In the second, applications run blindly on top of generic, “least common denominator” hardware built for nothing in particular. The first option revitalizes the semiconductor industry and opens the door to a new wave of electronic innovation. The second leads to economic and technological stagnation.
One key to enabling the first scenario is lowering the cost of new designs. A world in which new SoC designs cost $100 million and take two years to implement will not produce much hardware innovation. System-level design technologies can help by lowering hardware development costs, allowing early software development, and reducing time to market. The best possible result is an application-rich world in which hardware innovation really does matter.
–Richard Goering is manager of technical communications at Cadence Design Systems, and the author of the Cadence Industry Insights blog.