Not long ago, I overheard a thought-provoking exchange related to the demise of a beloved cultural icon. This conversation gave me pause to consider what intriguing new application is coming next that will displace another symbol of popular culture, much like what’s happened to local book and video stores or hard-wired landline phones. I’m sure some clever entrepreneur is already designing an enabling technology to open a new world for us, beyond our current habits. Gone may be a beloved local store or pink Princess phones, but consider the access to a variety of new adventures offered by these future innovations.
Ah, but these game-changing technologies don’t exist in a vacuum. They rely on innovations occurring at all levels of development. The next iconic technology will require more than just sophisticated application software and fast internet connections. It will also require a robust infrastructure able to crunch through and process, compress, and transport the massive amounts of data. That infrastructure will be built on the next generation of multicore (or manycore) SoC devices executing previously unimaginable amounts of embedded software. And these SoC devices will in turn be designed and verified using the latest and greatest EDA technologies, most notably, emulation.
To support the next big idea, every EDA tool requires innovation, but this is especially true for emulation, given its growing prominence in the process of SoC realization. As I’ve written in the past, emulation has become mandatory for the verification of complex chips. Thus, emulators can’t just keep up with the requirements of SoC development; they must outpace them. SoC design sizes of 20- or 30-million gates are common today, but the emulator for tomorrow must be able to cross the billion-gate threshold. Similarly, the 500kHz to 1MHz performance that emulators have traditionally supplied simply won’t cut it for next-generation SoC designs. You need an order of magnitude performance boost if you are going to boot an OS, transcode high-definition video, or process multiple pages of scanned images.
Innovation isn’t always about size or speed. Sometimes it’s about breaking down barriers and improving accessibility. Design teams around the globe have concluded that an emulator can accelerate the verification process across the entire SoC project. To support this surge in adoption, emulators must evolve into a cost-effective and flexible solution that supports both enterprise- and desktop-level usage.
I’m still pondering what the next big idea will be, all-the-while knowing that innovations in emulation will play an important supporting role. I welcome your thoughts.