EdgeCortix Launches SAKURA-II: AI Edge Thrust

 

EdgeCortix® Inc. has unveiled its SAKURA-II Edge AI accelerator, powered by the second generation Dynamic Neural Accelerator (DNA) architecture. Engineered for energy-efficient AI processing at the edge, it excels in handling complex Generative AI tasks like Large Language Models (LLMs) and Large Vision Models (LVMs). With up to 60 TOPS of 8-bit integer performance and 30 TFLOPS of 16-bit brain floating-point operations per second, SAKURA-II offers low latency, high accuracy, and compact form factors, making it ideal for diverse industries including manufacturing, robotics, and telecommunications.

Featuring the DNA-II engine, SAKURA-II enables real-time processing of multiple deep neural network models with low latency. The platform is complemented by the MERA software suite, facilitating seamless integration with leading development frameworks and model libraries like PyTorch and Hugging Face Optimum. SAKURA-II boasts benefits such as optimized memory bandwidth, real-time data streaming, mixed-precision support, and efficient power management.

Offered in various configurations including stand-alone devices, M.2 modules, and PCIe cards, SAKURA-II is available for reservation now, with delivery expected in the second half of 2024. CEO Sakyasingha Dasgupta emphasized its pivotal role in advancing Generative AI solutions at the edge, offering superior performance per watt and deployment flexibility.

More Information: https://www.techdogs.com/tech-news/business-wire/edgecortix-launches-sakura-ii-platform-to-power-the-next-wave-of-generative-ai-at-the-edge