사물인터넷(IoT), 산업 자동화, 스마트 로보틱스의 발전과 함께 의료 영상 솔루션에서도 엣지 AI가 확대됨에 따라 전력 효율성과 열 관리를 지원하는 애플리케이션의 개발이 그 어느 때보다 복잡해졌다.
Application-specific & integrated HW/SW technology stack
As edge AI expands into medical imaging solutions, along with advances in the Internet of Things (IoT), industrial automation, and smart robotics, developing applications that support power efficiency and thermal management has become more complex than ever.
Microchip Technology Inc. (NASDAQ: MCHP) today announced the availability of its PolarFire FPGA and SoC solution stack for smart robotics and medical imaging that accelerates product development cycles and simplifies complex development processes.
The newly launched products are based on Microchip’s existing smart embedded vision and industrial edge and intelligent edge communication stacks, and consist of: AI-enabled 4K60 computer vision firmware; IP cores; a variety of ready-to-use sensors; camera interfaces; and integrated hardware supporting high-speed Ethernet protocols.
Real-time ROS-2 compatible cores facilitate robotics tasks such as perception and coordinate transformation. It explained that the stack solution provides an industrial high-precision time networking protocol for OPC/UA, commonly used in industrial automation, rich operating system support, and asymmetric processing capabilities commonly used in industrial automation.
The software design kit (SDK) supports various development environments centered on C/C++, RTL, and machine learning frameworks, and supports customization through tools such as SmartHLS™ IDE, VectorBlox™ Accelerator SDK, and Libero SoC Design Suite with IEC61503 SIL 3 functional safety certification.
The company offers hardware and software solutions that combine mid-range PolarFire FPGAs and PolarFire SoC FPGAs, and emphasizes that cybersecurity protection features enable system designers to create creative designs in medical imaging and robotics applications.
“Our customers require unprecedented compute performance in a tiny physical footprint that is highly vulnerable to thermal management and cybersecurity threats,” said Shakeel Peera, vice president of Microchip’s FPGA strategic marketing division. “They also have an urgent need to drive innovations for enhanced security and functional safety in AI-based industrial automation and portable medical imaging.”
“To that end, Microchip is providing developers across all segments with the ability to leverage power-efficient hardware and customized solution stacks to accelerate the adoption of intelligent medical imaging and autonomous robotics,” he continued.
Meanwhile, the Microchip solution stack includes the PolarFire FPG, which works with the recently announced NVIDIA Holoscan sensor processing platform.It has an Ethernet sensor bridge and has the ability to transmit real-time sensor data to NVIDIA Holoscan and NVIDIA IGX, NVIDIA Jetson platforms, which are used for edge AI and robotics. This enables new edge-to-cloud applications and AI/ML inference.