New HSA Foundation specification brings us closer to heterogeneous computing

Share on linkedin
Share on twitter
Share on facebook
Share on reddit
Share on digg
Share on email

When the HSA Foundation was officially introduced back in 2012, I remember being on a conference call with our EVP of marketing Tony King-Smith and Patrick Moorhead*, founder and president of Moor Insights & Strategy.

Patrick was working on a whitepaper (you can read it here) and was talking to the founding members to get more background information on how and why the foundation had been created. Right from the start, one of interesting things that Patrick pointed out was the wave of positive feedback that was pouring in from everyone involved in computing, whether it was mobile and embedded, desktop or HPC.

The buzz in the air was real and palpable.

The HSA Foundation united AMD, ARM, Imagination, MediaTek, Qualcomm, Samsung and Texas Instruments in a quest to build interoperable hardware and software that would address the high-efficiency requirements of next-generation platforms.

HSA FoundationThe founding members of the HSA Foundation

Yesterday the HSA Foundation launched version 1.0 of the specification during an event held at the Fairmont Hotel in San Jose, California (read the press release here). All board members were there to discuss software, the ecosystem and applications in the mobile, desktop PC and high-performance computing.

HSA Foundation - Launch eventAll founding members of the HSA Foundation participated in a panel discussion

To accompany the launch, I have produced some examples of how system designers can use our processor IP to implement power-efficient heterogeneous platforms:

  • MIPS I6400 CPU: This new 64-bit CPU is designed for true heterogeneous processing for mobile and embedded applications; at the CPU level, it scales from one to four threads – and system architects can put up to six CPUs in a cluster. Furthermore, an I6400 processor is designed to scale up to 64 clusters, extending to over 1500 threads. Every CPU inside the cluster can be managed individually, allowing for fine-grain control of frequency and power.
  • All PowerVR Rogue GPUs feature a scalar SIMD architecture that treats general purpose compute as a first class citizen, while not forgetting what makes a shader core great for graphics. The latest PowerVR Series7XT GPUs feature a scalable 64-bit floating point co-processor per cluster for customers who want to use our IP in HPC applications. We’ve also made several GPU compute setup and cache throughput improvements resulting in up to 300% better parallel processing performance.
  • Ensigma RPUs continue the trend of moving more processing on chip. Designed for full system integration, these radio processors include a powerful DSP architecture that is capable of handling multiple wireless communications standards on a single chip, including all digital TV and radio formats as well as 802.11 Wi-Fi and Bluetooth.

One superb illustration of how these different processors come together in a true heterogeneous platform is our PowerVR Vision IP platform. Suitable for applications such as mobile and wearable, automotive (ADAS), surveillance or augmented reality, this platform includes multiple processing units, all working together efficiently to handle multimedia-related tasks.

PowerVR vision platformOur PowerVR Vision IP platform is an example of a heterogeneous system

For example, the PowerVR ISP can send metadata to the CPU, video encoder or the graphics processor to speed up certain tasks; developers can also distribute computer vision algorithms more efficiently across several units (e.g. heavy parallel-oriented code can run on the GPU, while other optional algorithms can be executed on the CPU SIMD engine).

On the software side, we’ve recently introduced the PowerVR imaging framework for Android. This includes a set of extensions to the OpenCL and EGL APIs that enable efficient zero-copy sampling of YUV and RGB camera data. The framework also includes low-level functions for integrating these extensions within the camera Hardware Abstraction Layer (HAL).

If you want to know more about the HSA Foundation, you can visit their website. Make sure you also follow us on Twitter (@ImaginationTech, @MIPSGuru, @PowerVRInsider, @GPUcompute) for the latest news and announcements from Imagination.

Editor’s note

* Patrick Moorhead is currently ranked one of the top technology analysts in the world and has extensive experience in covering the computing market, comprising personal computing, graphics or servers. You can read more about him here and follow him on Twitter (@PatrickMoorhead).

Alex Voica

Alex Voica

Before deciding to pursue his dream of working in technology marketing, Alexandru held various engineering roles at leading semiconductor companies in Europe. His background also includes research in computer graphics and VR at the School of Advanced Studies Sant'Anna in Pisa. You can follow him on Twitter @alexvoica.

Please leave a comment below

Comment policy: We love comments and appreciate the time that readers spend to share ideas and give feedback. However, all comments are manually moderated and those deemed to be spam or solely promotional will be deleted. We respect your privacy and will not publish your personal details.

Blog Contact

If you have any enquiries regarding any of our blog posts, please contact:

United Kingdom
Tel: +44 (0)1923 260 511

Search by Tag

Search by Author

Related blog articles

bseries imgic technology

Back in the high-performance game

My first encounter with the PowerVR GPU was helping the then VideoLogic launch boards for Matrox in Europe. Not long after I joined the company, working on the rebrand to Imagination Technologies and promoting both our own VideoLogic-branded boards and those of our partners using ST’s Kyro processors. There were tens of board partners but only for one brief moment did we have two partners in the desktop space: NEC and ST.

Read More »
pvrtune complete

What is PVRTune Complete?

PVR Tune Complete highlights exactly what the application is doing at the GPU level, helping to identify any bottlenecks in the compute stage, the renderer, and the tiler.

Read More »


Sign up to receive the latest news and product updates from Imagination straight to your inbox.