With more and more products incorporating computer vision functionality, semiconductor vendors are increasingly designing specialized processors for vision applications. For product developers, this means that identifying the best processor (whether a chip for use in a system design, or an IP core for use in an SoC) is challenging—for several reasons.

CEVA's DSP cores have long been a fixture in many mobile phone cellular modems, where they handle the tranceiver's digital signal processing functions, working alongside function-specific hardware accelerators. However, the physical layer coordination between the transmit and receiver subsystems, along with overall modem control and other housekeeping tasks, has long been the bailiwick of a real-time CPU core from ARM, Imagination Technologies or another supplier.

ADAS (advanced driver assistance systems) are rapidly being incorporated into automobiles and other vehicles, as products unveiled at January's North American International Auto Show in Detroit, Michigan made clear. Just a few short years ago, passive collision warning and active collision avoidance features were restricted to luxury models from luxury manufacturers. Now, even mass-market car suppliers are incorporating ADAS capabilities into their high-end and mainstream vehicles.

It's no secret that sensors are proliferating. Our smartphones, for example, contain accelerometers, magnetometers, ambient light sensors, microphones – over a dozen distinct types of sensors. A modern automobile contains roughly 200 sensors.

As computer vision is deployed into a variety of new applications, driven by the emergence of powerful, low-cost, and energy-efficient processors, companies need to find ways to squeeze demanding vision processing algorithms into size-, weight-, power, and cost-constrained systems. Fortunately for its clients, BDTI's skill in benchmarking has armed it with unique skill in optimizing software to best exploit processor capabilities.

Just prior to the 2014 Consumer Electronics Show, Imagination Technologies unveiled its first computer vision processor offering with the announcement of the Raptor core architecture, the first product iteration of which was relea

Automobile-based processing intelligence, both in the form of fully autonomous vehicles and more modest ADAS (Advanced Driver Assistance Systems), garnered exclusive billing in NVIDIA's keynote and booth at this year's Consumer Electronics Show, held last month in Las Vegas, Nevada.

Today, billions of hours of video are collected each year, but most of it is never used, because we don't have a practical way to extract actionable information from it. A new generation of computer vision solutions, powered by deep neural networks, will soon change this, unleashing the tremendous value that's currently locked away in our video files.

As embedded processors and applications become increasingly complex, good benchmarks are more important than ever. System designers need good benchmarks to judge whether a processor will meet the needs of their applications, and to make accurate comparisons among processors. Processor developers need good benchmarks to assess how their processors stack up against the competition, and to prove their processors' capabilities to customers.

But what exactly comprises a good benchmark?

Hear the words "high volume DSP market" and you might automatically think of "mobile phones". And you'd be right; recent estimates peg quarterly worldwide mobile phone shipments approaching half a trillion units, with smartphones (which often contain multiple DSP cores) representing three-quarters of that amount.