02-02-2016 | | By Paul Whytock
What is said to be the world’s first hardware virtualised GPU products have been announced by US semiconductor company Advanced Micro Devices (AMD).
Called the AMD FirePro S-Series GPUs with multiuser GPU (MxGPU) technology, this hardware-virtualised GPU architecture is claimed by AMD to provide an innovative product relative to virtualisation processes.
In those sort of systems, components like the CPU, network controller and storage devices are being virtualised in hardware to deliver improved user experiences but prior to this announcement from AMD the GPU was not hardware virtualised.
What does this mean? AMD MxGPU technology, based on SR-IOV (Single Root I/O Virtualisation), a PCI Express standard, delivers precise hardware GPU scheduling logic to the user.
In addition it maintains the data integrity of Virtualised Machines (VM) and their application data through hardware-enforced memory isolation logic preventing one VM from being able to access another VM’s data. It also exposes all the graphics functionality of the GPU to applications which facilitates complete virtualisation support for not only graphics APIs like DirectX and OpenGL but also GPU compute APIs like OpenCL.
The company has also said that its new AMD FirePro S7150 and AMD FirePro S7150 x2 server graphics cards will combine with OEM offerings to create high-performance virtual workstations and address IT needs of simple installation and operation, and good data security.
From the perspective of IT budgets, these can realise support for up to 16 simultaneous users with a single AMD FirePro S7150 GPU card which features 8GB of GDDR5 memory, while up to twice as many simultaneous users (32 in total) can be supported by a single AMD FirePro S7150 x2 card which includes a total of 16 GB of GDDR5 memory (8GB per GPU).
Both models feature 256-bit memory bandwidth.
Based on AMD’s Graphics Core Next (GCN) architecture to optimise performance, the AMD FirePro S7150 and S7150 x2 server GPUs feature; Multiuser GPU (MxGPU) technology, GDDR5 GPU Memory to help accelerate applications and Error Correcting Code memory to ensure the accuracy of computations.