Imagination Technologies' E-Series GPUs Feature a New IP for AI Loads
Colin Smith — May 8, 2025 — Tech
References: imaginationtech & venturebeat
Imagination Technologies’ E-Series GPUs have been introduced as a new GPU IP designed to address both advanced graphics and artificial intelligence (AI) processing at the edge. The chips leverage a highly efficient parallel processing architecture to deliver scalable performance ranging from 2 to 200 TOPS (INT8/FP8) for AI workloads. Two key technologies—Neural Cores and Burst Processors—underpin the design: Neural Cores accelerate AI and compute tasks, while the Burst Processors improve average power efficiency by 35% for edge applications. The architecture also supports advanced graphics features, including ray tracing, thereby positioning the E-Series as a versatile solution for both high-performance graphics and low precision AI operations.
In addition to its hardware innovations, the E-Series GPUs offer significant programmability and integration benefits. The design supports a variety of APIs, such as OpenCL, and is compatible with developer tools including oneAPI, Apache TVM, and LiteRT, enabling efficient porting and optimization of workloads. The local memory architecture further reduces the power and performance costs associated with external memory access, which is critical for edge computing environments. Moreover, enhanced virtualization capabilities allow for up to sixteen hardware-backed, zero-overhead virtual machines, adding flexibility for a range of applications including mobile devices, industrial vision systems, and automotive systems.
Image Credit: Imagination Technology
In addition to its hardware innovations, the E-Series GPUs offer significant programmability and integration benefits. The design supports a variety of APIs, such as OpenCL, and is compatible with developer tools including oneAPI, Apache TVM, and LiteRT, enabling efficient porting and optimization of workloads. The local memory architecture further reduces the power and performance costs associated with external memory access, which is critical for edge computing environments. Moreover, enhanced virtualization capabilities allow for up to sixteen hardware-backed, zero-overhead virtual machines, adding flexibility for a range of applications including mobile devices, industrial vision systems, and automotive systems.
Image Credit: Imagination Technology
Trend Themes
1. Edge AI Processing - Enhancing edge AI processing through highly efficient parallel architecture provides a scalable solution for computing demands from 2 to 200 TOPS.
2. Power-efficient AI Hardware - The integration of Burst Processors significantly boosts power efficiency by 35%, paving the way for more sustainable edge computing applications.
3. Advanced Graphic Integration - The E-Series GPUs' support for advanced graphic features like ray tracing offers innovative prospects for high-performance visual applications in edge devices.
Industry Implications
1. Mobile Device Manufacturing - Enabling the deployment of versatile and efficient GPUs allows for improved AI and graphics capabilities in next-generation mobile devices.
2. Industrial Automation - Implementing GPUs with enhanced virtualization supports multiple applications, including smart industrial vision systems.
3. Automotive Technology - The GPUs' capability for offering both high-performance graphics and low precision AI tasks can revolutionize automotive systems with advanced driver-assistance features.
4.2
Score
Popularity
Activity
Freshness
