Imagination Technologies has launched its E-Series GPU IP, designed to accelerate both graphics and AI at the edge. This new GPU architecture offers significant performance gains, scaling from 2 to 200 TOPS INT8/FP8 for AI workloads while providing exceptional graphics performance. The E-Series GPUs feature Neural Cores that deliver up to 200 TOPS, accelerating AI and compute workloads, and supporting various AI number formats for flexibility and performance efficiency.
Additionally, the E-Series GPUs include Burst Processors, which improve power efficiency by 35% for AI workloads, games, and user interfaces. These GPUs are programmable and future-proof, designed to adapt to evolving AI, compute, and graphics workloads, with support for popular APIs like OpenCL and frameworks like oneAPI and Apache TVM.
The E-Series GPUs are expected to be used in various applications, including edge AI systems, graphics-rich applications, natural language processing on smartphones, industrial computer vision, and vehicle autonomy. The first E-Series GPU IP is set to be available in Autumn 2025, with automotive, consumer, desktop, and mobile variations in development. This launch positions Imagination Technologies as a key player in the edge AI and graphics market.
