Tim Vehling, Executive VP Global Sales & Business Development at EdgeCortix, discusses the company's focus on delivering near-cloud-level AI performance at the edge with high energy efficiency and reduced operating costs. He highlights the rapid growth in AI model size and complexity, emphasizing the challenges of deploying generative AI on edge devices due to increasing compute and memory demands.
EdgeCortix's SAKURA-II platform addresses these challenges by offering a multimodal, high-efficiency AI accelerator with robust DRAM bandwidth and capacity, making it suitable for running advanced models, such as vision transformers and large language models, on embedded systems. Tim concludes by outlining their full-stack solution, recent innovations such as running generative AI on Raspberry Pi, and the company's roadmap for even more powerful next-generation edge AI architectures.
EdgeCortix Edge AI Platform
Full-Stack Solution: Integrated Platform Increases Ecosystem Value Over Time
MERA Compiler and Framework
Industry first software platform enabling AI inference across heterogeneous systems
