Data Centre &

Desktop

Enabling high-performance graphics with low power

Step up to the future of acceleration

With more than 30 years of experience, we have led the way in GPU compute, Neural Network and Communications acceleration, time and again delivering the highest performance density per mm2 in the market.

4x

Multi-Core Performance

30%

Lower Power

2.5x

Fill Rate

The XT GPU family

Enable your high-performance GPU solution with the fastest GPU IP ever created. Our multi-core architecture enables our customers to link together multiple GPU cores to achieve the highest levels of performance with new configurations to match existing and emerging computing trends. Whether you need a compute-focused core to enable cloud computing or all-round GPU performance to deliver mobile or desktop games through the cloud, we offer an optimised GPU IP solution ideally suited to your needs.

Download multi-core overview

Fantasy one, Innosilicon’s first performance server grade graphics card, integrates IMG BXT multi-core GPU architecture into its highly innovative SoC.

Find out more

“By integrating IMG B-Series GPU technology, we were able to create an incredibly power-efficient solution, that delivers 4K performance in a PCI-E Gen 4 GPU”

Roger Mao

Executive Vice President

The new desktop

The Imagination GPU density advantage means we can deliver superfast desktop performance while maintaining the PPA advantages that have made us the leaders in mobile and automotive GPU.

The gaming cloud

Cloud-based gaming is expanding the range of options for gamers, but power consumption and heat management are key challenges faced by server farms. Our high efficiency GPUs deliver high performance with the lowest power demand and thermal footprint per watt. Our advanced virtualization features make our GPUs ideal for running games 1:1 or 1:many for the most efficient cloud gaming solutions.

Time-sensitive networking

Networking at the Edge is becoming more and more critical to delivering quality of experience and service. This means that the functionality of the switching elements need to be both deterministic as well as low in terms of latency and other technologies to increase the reliability of the packets for industrial and communication networks.

Inside the data center the distributed computing architectures require the same capabilities across the network interfacing elements of the compute, storage and networking elements.

OpenGL®, Vulkan® and beyond

We have the software you need for success including OpenGL, OpenGL® ES and Vulkan® API support and advanced developer tools such as PowerVR Carbon, our powerful API tracing and debugging tool and PVRTune, our industry-leading profiling and analysis utility for graphics applications.

Our developer community is vast, and with the Imagination Developer Portal, you can find resources and advice on the best way to utilize the feature-rich PowerVR Tools suite.

Get started

Ray tracing: the next level

Imagination’s unique ray tracing architecture contains patented specialist hardware blocks that enable either faster full ray tracing or an efficient hybrid rendering approach that combines traditional rasterisation techniques with ray tracing for realistic lighting, reflections and shadows even on low power mobile devices.

Our highly scalable ray tracing solution enables cutting-edge cloud gaming, delivering stunning visual-effects quality combined with low power consumption delivering the robustness demanded by high-volume cloud gaming environments.

Contact us

Get in touch

Get in touch and sharpen your competitive edge with our high performance graphics.

Loading Form…