E-Series GPU IP features a deeply integrated Matrix Accelerator and support for low-precision number formats, ensuring fast and flexible acceleration for cutting-edge gaming and AI workloads.
GRAPHICS
NEURAL RENDERING
COMPUTE & AI
Imagination E-Series GPU IP unifies graphics, compute and AI on one processor, with one software stack.
Matrix acceleration units deep inside the GPU’s shading clusters deliver efficient acceleration for popular but compute-heavy edge workloads like computer vision, generative AI and neural rendering.
The cost of developing advanced chips is enormous. Every mm² of silicon has to deliver value throughout the project’s lifecycle.
A general-purpose, GPU-based accelerator that can render a user interface, support an overloaded NPU and be trusted to run new models as they appear adds considerable value to a modern, edge AI system.
Download the architecture overview
The Architecture of an E-Series Core
Understand how one GPU IP can handle graphics, compute and AI tasks. The single-core EXD-64-2048 GPU shown on the left delivers 64GTexel/s, 2 TFLOPS FP32 and 32 TOPS INT8 @ 1GHz.
The firmware processor is in charge of scheduling workloads across the GPU, communicating with third party accelerators and handling errors. It operates independently of the CPU delivering superior performance.
The USC is the programmable compute engine at the heart of every Imagination GPU. Inside each USC, an ultra-wide 128-lane ALU delivers massive parallel processing throughput, enabling efficient execution of graphics shaders, neural networks and general-purpose compute workloads on the same architecture.
Matmul acceleration with support for high and low precision number formats (FP32, FP16, BF16, INT8, FP8, MXFP8, FP4, MXFP4) is embedded deeply within the Imagination GPU’s ALU’s, sharing the same register storage for maximum data efficiency and power consumption.
Customers can scale E-Series up to cloud levels of performance using Imagination’s decentralised, chiplet-friendly multi-core technology.
A quad-core E-Series GPU running at 1.6GHz reaches nearly 200 TOPS INT8.
Available E-Series Configurations
View All ProductsE-SERIES: GRAPHICS EXCELLENCE
+39%
power efficiency
+54%
real-world gaming uplift
60FPS
PC gaming
MORE GAMING PER UNIT OF COMPUTE
E-Series combines Burst Processors with a re-balanced Scalable Processing Unit in order to deliver massive gains on gaming performance and efficiency compared to D-Series.
Scaling up to 256GTexel/s, E-Series can run popular AAA games at over 60FPS while support for DirectX 12 FL11_0 enables devices to run Windows 11 and even more titles.
E-SERIES: NEURAL RENDERING
The deep integration of the matrix acceleration inside the E-Series ALU is ideal for neural rendering applications. With AI and rendering units tightly coupled, we reduce data travel, increase power efficiency and sustain performance.
NEURAL SUPER RESOLUTION
An extremely efficient single-pass temporal upscaling solution deeply optimised for Imagination GPUs and accessible via industry-standard APIs.
Neural Super Resolution can deliver a significant increase in game frame rates and is particularly useful for delivering complex effects on constrained devices.
2.3ms
latency
54%
lower memory bandwidth consumption
65%
model weight loss from self-compression
E-SERIES: AI ACCELERATION
4.4x
Conv2d speed-up
4.8x
GEMM speed-up
89%
GPU utilisation
POWERING THE EDGE AI EXPERIENCE
Deeply integrated matrix acceleration, support for low precision number formats and optimised kernel libraries combine on general-purpose silicon to create a powerful edge AI solution that can confidently tackle the operators of the present and the future.
For Qwen 3.5 4B, a quad-core E-Series GPU at 1.5GHz can achieve time to first token of 0.2 seconds and decode throughput of 150 tokens per second on a representative workload.
To make life easy for developers, an optimised backend for Llama.cpp is being upstreamed, with native PyTorch and ONNX Runtime to follow.
Learn about running AI on E-Series
WHY CHOOSE E-SERIES FOR EDGE AI?
FLEXIBILITY
E-Series gives edge AI systems an extra option for executing compute-heavy and AI workloads. Busy NPUs can offload tasks onto the GPU which has multiple features for handling AI and graphics workloads concurrently.
FUTURE-PROOF
New AI models emerge each month, while hardware needs to stay in the market for years. Fixed function accelerators struggle to run new operators. A general-purpose, programmable accelerator protects your SoC against a software-defined future.
ACCESSIBILITY
E-Series provides developers with a familiar GPU programming model and leverages industry-standard APIs like Vulkan and OpenCL. The software stack integrates into popular AI frameworks for easy porting, with libraries and compilers to deliver maximum GPU utilisation.
Upcoming Webinars From Imagination
The Convergence of Edge GPUs
20 October 2026 | Hosted by SemiWiki
Upcoming Webinars From Imagination
Porting, Analysing & Optimising AI Models onto Edge GPUs
11 November 2026 | Hosted by SemiWiki
Upcoming Webinars From Imagination
Creating A Hardware-Optimised Neural Rendering Solution
8 December 2026 | Hosted by SemiWiki
OUR EXPERTS ON E-SERIES GPU IP
The GPU Inside 12Bn+ Devices
Imagination has spent more than thirty years optimising our innovative GPU architecture for gaming, user interfaces and compute-heavy applications. We license the architecture as a soft GPU IP, our licensees build it into their silicon, and that silicon ships in billions of cars, phones, wearables and PCs worldwide.
DESKTOP
> Desktop Productivity
> PC Gaming
> Professional Rendering
> Cloud Gaming
AUTOMOTIVE
> In-vehicle Infotainment
> Cockpit Display
> Sensor Fusion
> ADAS Functions
MOBILE
> Graphical User Interfaces
> Mobile Gaming
> Language Models
> Neural Rendering
CONSUMER & IoT
> Graphical User Interfaces
> Voice & Gesture
> Data Pre-Processing
> Computer Vision
A NEW ERA OF ON-DEVICE AI AND GRAPHICS
“On-device AI is evolving rapidly, but edge AI system designers still face challenges in balancing performance and efficiency with flexibility… Imagination has leveraged its long-held experience developing power-efficient GPUs and evolved them to flexibly support both graphics and AI workloads for on-device AI. The E-Series combines the programmability of GPUs with a boost in AI performance to create a compelling solution for on-device AI system designers.”
Research Director
INTERESTED? CONTACT US NOW
Want to learn more about E-Series and how we can help elevate your business to new heights?
Complete our contact us form, and let one of our experts guide you to success.
