The device is programmable and is intended for future edge applications including graphics, desktop applications, natural language processing on smartphones, industrial computer vision, and vehicle autonomy.
Two new technologies underpin the E-Series: Neural Cores, scaling up to 200 TOPS (INT8/FP8) that provide significant acceleration for AI and compute workloads and Burst Processors that deliver a 35% improvement in average power efficiency for edge applications.
“On-device AI is evolving rapidly, but edge AI system designers still face challenges in balancing performance and efficiency with flexibility,” said Phil Solis, Research Director at IDC. “The E-Series combines the programmability of GPUs with a boost in AI performance for on-device AI system designers.”
The E-Series offers advanced graphics capabilities but also adds deeply integrated acceleration for power-efficient, low precision AI operations into every GPU core. This creates the compute-dense E-Series Neural Cores that scale up to 200 TOPS INT8 and deliver up to 400% the AI performance of the previous D-Series. The Neural Cores support a wide range of AI number formats, enabling developers to design networks that meet a broad spectrum of demands. In addition, an AI-friendly memory architecture prioritises local memory for compute, greatly reducing the power and performance costs of going to external memory.
The E-Series Neural Cores align with the broader GPU and heterogeneous computing software ecosystem by integrating AI acceleration across the GPU. Their capabilities can be unlocked by popular APIs, like OpenCL, and developers can move their workloads onto the Neural Cores using open standards and tools.
The E-Series’ new Burst Processors technology enhances power efficiency and does this by reducing pipeline depth and minimising data movement within the GPU.
The series also enhances the multitasking capabilities of previous generations by doubling the number of hardware-backed, zero-overhead virtual machines supported by Imagination GPUs to sixteen, with sophisticated QoS support. Multi-core variants of E-Series GPUs can leverage additional cores for either extra performance or enhanced flexibility.
“The E-Series will place the GPU at the centre of both graphics and edge AI systems,” said Tim Mamtora, Chief of Innovation and Engineering, Imagination. “For system designers needing to run both graphics and compute workloads, an E-Series GPU is a versatile solution that eliminates the need for additional vector-based or fixed function AI solutions, giving future-proof flexibility while saving on total system design costs.”
The first E-Series GPU IP is available in Autumn 2025 and has already been licensed.