AI &
COMPUTE
We deliver scalable and power-efficient AI performance across a wide range of devices, from edge computing to data centres.
Adding Flexibility into Intelligent Systems
Whether it’s in your hand, behind your TV, or in factory robots, Imagination’s GPU IP delivers performance and efficiency. As AI becomes essential everywhere, system designers leverage Imagination’s ultra-parallel GPU architecture to accelerate AI models on power-constrained devices.
Imagination GPUs are highly flexible, running AI and graphics workloads simultaneously. Their programmability is enhanced by a developer-friendly compute software stack, helping devices stay ahead in the rapidly evolving AI landscape.
Our latest AI updates
Super-fast compute scheduling
OpenCL Compute Libraries
Expanded number format support
Lots of local memory
Why Imagination for AI?
Flexible graphics & compute processing
Our GPUs’ asynchronous processing of different task types – like rendering the smartphone’s UI and accelerating an LLM – makes them a very flexible silicon investment. They also come with hardware-based virtualisation for low-overhead GPU multitasking.
Efficient by design
Its tile-based approach to computing keeps as much data local to the GPU as possible and is just as applicable to compute workloads as to graphics.
Masters at multitasking
Their support for asynchronous processing of different task types – like rendering the smartphone’s UI and accelerating an LLM – makes the m a very flexible silicon investment.
All Imagination GPUs come with hardware-based virtualisation for fully secure GPU multitasking.
Imagination GPUs for AI
D-Series
Imagination’s D-Series takes GPU density, efficiency and performance to a new level and adds innovative extras that help our partners succeed in competitive markets – whether that’s enabling automotive chip designers to achieve ASIL-B functional safety without overhead, or halving the area penalty of mobile ray tracing.
C-Series
Imagination's efficient GPUs improve user experiences across a wide range of consumer devices. IMG CXM is Imagination’s latest, highly efficient GPU, available in a variety of configurations to ensure the best fit for your project.
B-Series
B-Series covers all of Imagination's markets, scaling from low-area configurations for set-top boxes right the way through to high-performance solutions for desktop. B-Series introduces Imagination's innovative multi-core technology for boosting performance or adding extra multitasking flexibility to the GPU.
A-Series
IMG A-Series contains Imagination's smallest GPU IP and is perfect for industrial settings, consumer devices and entry-level mobile. Optimised for power efficiency, A-Series delivers consistent, sustainable frame rates without clock throttling.
Key Markets for AI
Automotive
The journey from driver assistance to automated driving is centred on the evolution of the capabilities of AI – both on the hardware and the software side. From driver monitoring to path management, today’s cars are growing in intelligence supported by flexible, scalable and programmable accelerators from Imagination.
Desktop
Generative AI is transforming business productivity and creative processes. A new era of AI laptops are emerging with new levels of compute power – but energy efficiency is still an essential feature. Imagination’s high-performance GPUs with DirectX are perfect for delivering AI-on-the-move.
Mobile
Whether it’s removing an unwanted person from a selfie, adjusting the resolution of a favourite picture or delivering a high-quality voice-based user interface, AI on mobile is here to stay and requires exceptionally capable processors that can handle complex AI algorithms with a limited power budget. Imagination’s energy efficient processors are synonymous with smartphones and our technology can support OEMs find differentiation through advanced AI features.
Consumer
On popular consumer devices from DTV to the Smart Home Hub, edge-based AI workloads are being introduced so devices can stand out from the competition, from gesture recognition to natural language processing. But this hardware market is cost-sensitive and needs to deliver new features within a constrained silicon area. Imagination’s CPU and GPU IP for consumer devices pack a huge amount of performance and AI features into a small package.
DXTP GPU IP boosts performance while extending battery life on power-constrained devices
Read the full press release“Zelos is integrating the advanced, power-efficient compute of Imagination DXTP into our upcoming chip. Imagination’s GPUs combine the performance of ultra-parallel processing with the flexibility that comes from a highly programmable architecture, making them the ideal platform for accelerating our AI models.”
Technical Director
Frequently asked questions
AI Accelerator chips come under many names such as Neural Network Accelerators (NNAs), Neural Processing Units (NPUs) and Machine Learning Engines. They are specialised processors designed to handle the complex computations required for artificial intelligence (AI) applications. Some examples of products that use AI accelerator chips include:
- Smartphones: Many high-end smartphones, such as the iPhone 12 and Samsung Galaxy S21, use AI accelerator chips to power features such as facial recognition, voice recognition, and augmented reality.
- Smart home devices: Smart home devices such as Amazon Echo and Google Nest Mini 2nd gen use AI accelerator chips to process voice commands and provide intelligent responses.
- Self-driving cars: Autonomous vehicles use AI accelerator chips to process sensor data and make real-time decisions based on the surrounding environment. Read more about AI in self-driving cars.
AI processors offer several benefits over traditional processors, including:
Faster performance: AI processors are designed to handle the complex computations required for AI workloads, such as deep learning and machine learning, much faster than traditional processors. This allows for more efficient processing of large datasets and faster training of AI models.
Energy efficiency: AI processors are optimised for processing large amounts of data in parallel, which can be done more energy-efficiently than traditional processors. This means that AI workloads can be processed more quickly and with less energy consumption helping companies achieve net zero.
Improved accuracy: AI processors are designed to handle the specific computations required for AI workloads, which can lead to improved accuracy in AI models. This is especially important in applications such as image recognition or natural language processing, where accuracy is critical.
Scalability: AI processors can be scaled more easily than traditional processors, which allows for faster processing of larger datasets and more complex AI models. This makes it possible to train and deploy AI models more quickly and efficiently.
Specialised design: AI processors are designed specifically for AI workloads, which means they can perform computations that would be difficult or impossible for traditional processors. This opens up new possibilities for AI applications, such as real-time object detection or speech recognition.
Overall, the benefits of AI processors make them essential for many AI applications, from self-driving cars to voice assistants to medical diagnosis tools.
