AMD announces Heterogeneous Queuing tech
October 22, 2013 // 9:10 a.m.
AMD has announced the latest string in its Heterogeneous Systems Architecture (HSA) bow: Heterogeneous Queuing, a platform for making the full compute power of hybrid accelerated processing unit (APU)-like chips available to the system.
Part of the HSA platform formerly known as Fusion, hQ is an open standard with which AMD hopes to push its hybrid vision of computing whereby the distinction between central and graphics processors becomes increasingly blurred. Designed to work alongside other HSA technologies, including the hUMA unified memory system, hQ looks to eliminate bottlenecks when generating work for the GPU.
To explain: traditional queuing systems work by forcing the CPU to run through an operating system service and kernel-level driver when generating workloads for the GPU. This, AMD claims, introduces unwanted latency to the queue - and means that every time there's work for the GPU to carry out, the CPU has to get involved to kick things off.
The hQ paradigm, by contrast, upgrades the GPU from a mere adjunct to the CPU to a processor of equal status. In this design, a given application can generate tasks queues directly on the GPU without the CPU getting involved. Better still, the GPU can generate its own workload - AMD's example here is raytracing, where one GPU task may generate several more tasks in its execution and which would previously have needed the CPU and operating system to be involved in queuing said new tasks - and even push work into the CPU's task queue. Equally, the CPU can push work into the GPU task queue without operating system involvement - creating a bi-directional queueing system which dramatically reduces latency and allows applications to easily push jobs to whichever processor is most appropriate.
For AMD, hQ is an important move: the company's flagship processors have long lagged behind rival Intel's on pure-CPU performance. Where the company shines is in its hybrid CPU-GPU products, and while harnessing the total compute power of such chips isn't straightforward today the company clearly hopes that the HSA Foundation and technologies like hUMA and hQ will help with that - and help unlock the hidden performance potential in its chips.
AMD explained to bit-tech that it will be releasing the technology to its HSA Foundation partners first for Windows and then for Linux and other operating systems. The technology is also expected to be picked up by other HSA Foundation members, including ARM, Samsung, Qualcomm, Mediatek, Texas Instruments and Imagination Technology - pointing to a future where hQ-powered GPGPU technology is used to boost performance and battery life on mobile devices. It will, however, require a new generation of chips before making its way into commercial products.
The cross-vendor support will be helped, AMD explained, by the use of the HSA Architected Queuing Langage (AQL), which provides a standardised packet format for hQ - meaning that application developers won't need to worry about which vendor they're dealing with when building hQ support into their software.