Today, Ambarella, Inc, an AI vision company, announced that they are expanding their AI vision SoC portfolio with the new CV5S and CV52S security.
The new SoCs support 4K encoding and advanced AI processing in a single low-power design. It is based on the CVflow® architecture and advanced 5nm process technology which provides industry-leading edge AI SoC performance.
The CV5S is for security camera applications that require multiple sensors over a wide area for 360-degree coverage. It is designed for single-sensor security cameras that need to identify individuals and objects.
John Lorenz, senior technology & market analyst at Yole Développement, says “Security system designers desire higher resolutions, increasing channel counts, and ever faster and more ubiquitous AI capabilities.”
He further added, “Ambarella’s newest AI vision SoCs for security, the CV5S, and CV52S, are competitive solutions for meeting the growing demands of the security IC sector, which we forecast to exceed $4 billion by 2025, with two-thirds of that being chips with AI capabilities.”
CV5S SoC can also support multi-image camera designs while simultaneously processing and encode four imager channels. Besides, these SoCs double the encoding resolution and memory bandwidth than Ambarella’s prior generation.
CV52S SoC family supports 4K resolution at 60fps while providing 4x the AI computer vision performance, single-sensor security cameras, 2x the CPU performance, and 50% more memory bandwidth. Neural network (NN) performance enables more AI processing performance at the edge.
Jerome Gigot, senior director of marketing at Ambarella, said “The global security industry is rapidly moving to higher 4K resolutions while increasing AI algorithm capabilities, to achieve better recognition of people and objects, along with multiple images for a wider field of view and longer range.”
“Our new CV5S and CV52S security AI vision SoC families respectively support 4K multi-imager or high-frame-rate 4K single-image designs with high-performance edge AI processing, allowing the development of cameras that do not need to compromise between image resolution and AI processing accuracy.”
Other than security, other AIoT applications can also take advantage of the high resolution and advanced AI processing. For instance, smart cities can leverage great image resolution and high edge AI performance for improved traffic management, automated speed control, accident detection, and finding the location of missing vehicles.
Similarly, retailers can use this advanced AI resolution to better manage product placement, enhance warehouse product tracking, real-time line management, and bring actionable intelligence to stores.