Pioneering the future of verification
Experts gather to discuss the state-of-the-art and future directions for testing and verification
How can you bring about transformation in the field of engineering design automation? This was one of the central questions addressed by Kathryn Kranen, President & CEO of Jasper Design Automation, in her talk at the recent Haifa Verification Conference HVC 2011. The conference is focused on different types of verification, which is the art of making sure a system works as designed.
Kranen explained to the audience of over 100 verification experts from academia and industry that having innovation is important – but it's not enough. Her talk focused on the successful path of generating new verification technologies by gradually adjusting to the changing needs of the market.
"Technological innovations must move in lockstep with business innovations for successful market adoption," noted Kranen. "Each 'step' of industrial usage generates revenue—and real-world feedback—to fuel the next set of innovations." By mastering this execution model, she feels that organizations can bring about market revolutions in an incremental fashion.
The three-day HVC 2011 has already established itself as a high-level international conference. Recognized as a center of competence for verification, IBM Research – Haifa is a natural place for colleagues to meet, discuss research ideas, and spark new collaborations. Hosted by IBM and sponsored by big players like Cadence, Mentor Graphics, and Jasper Design Automation, the conference is a much needed gathering for experts from Israel and abroad. Participants included the sponsors themselves, Israeli startups, major manufacturers such as Tower and Intel, academics, and graduate students.
Heading into the great unknown?
Many of the lectures, including the talk by Klauss-Dieter Schubert from IBM Germany, touched on the fact that movement to multicore will likely prevail for a while. In recent years, there has been no real acceleration of processor clocks. Schubert's talk described why continuing with "business as usual" will fail going forward and then presented a number of scenarios for workload optimized systems to overcome these hurdles.
Dr. Onn Shehory, senior research at IBM and general chair of HVC 2011. "The new growth will be in the area of parallelism, multicore, and perhaps something new that is not yet known today." Shehory believes that verification and testing will gradually expand their focus on parallel cores, on cases of shared memory in multicores, and on concurrency."Although processor speeds used to increase every year, we are now getting closer to the physical limits of the silicon material and speedup is stagnating," stated
Engineered systems to the rescue
Another 'hot topic' both at the conference and in the industry is the growing complexity of today's systems. For example, looking at a car, we see an engineered system that includes physical components, computer hardware, and very complex software—which can run to over one million lines of code. As systems become more and more sophisticated, verifying single components is no longer sufficient. The design of the system itself must begin to incorporate ways to validate and test the system.
This emerging field is quickly gaining popularity in the post academic and business communities. This is especially important since some 50 percent of systems engineering cost is spent either performing Verification, Validation, and Testing (VVT) activities or correcting system defects during the development process or during product's lifetime. The HVC tutorial day included a session on VVT methodologies given by Dr. Avner Engel, author of the book "Verification, Validation and Testing of Engineered Systems" published by Wiley in 2010. Engel covered practical tips for developing a methodology and a process model for systems verification, validation, and testing.
"The verification, validation and testing of engineered systems is capturing attention in the field now and will most likely influence verification directions in the future," noted Amir Nahir, verification expert at IBM Research – Haifa, one of the HVC 2011 organizers, and chair of HVC 2012.
Looking to the future
Who will lead the change in direction for verification? "Future verification solutions will most likely evolve slowly to encompass new more complex systems and concurrency issues – but still remain based on the current core," explained Nahir. "New breakthrough solutions will need to evolve through the joint collaboration of different companies or research work—and this is exactly the kind of teaming we achieve through conferences such as HVC."