CHIP MAKER AMD has announced the imminent release of the Vulkan 1.0 specification, an API built from its Mantle GPU technology.
AMD will also release the beta version of a Vulkan-equipped software driver for its Radeon graphics cards designed for gaming and high-performance 3D rendering work.
Vulkan was formed from a non-profit consortium of companies called the Khronos Group, which sounds like something out of Doctor Who and includes graphics tech firms Nvidia and AMD, and offers a low-overhead API for GPU drivers.
This provides a way for graphics card drivers to reduce the amount of background processing power drawn from a CPU when interpreting the hardware demands required for running 3D applications.
Vulkan instead allows 3D graphics applications to have more direct control over the resources of a GPU, thereby cutting down on CPU overhead.
The API can also improve the spread of 3D processing work across multiple CPU cores. Again, this cuts down on the amount of work a CPU needs to do when running graphically intensive applications.
Vulkan is designed to complement the OpenGL graphics API, and AMD claims that it can access features in GPU hardware not normally available when using only OpenGL. It also supports Windows versions 7 to 10, as well as Linux.
Vulkan has also wormed its way into the Android ecosystem after Google announced that its mobile OS will support the API.
AMD claims that Vulkan 1.0, when used with its Radeon graphics cards, will help PC game developers remove software bottlenecks in graphics processing and suck more power from the GPUs.
AMD’s work on Vulkan as part of the Khronos Group indicates that the company is happy to play nice with others to produce new graphics tech.
“The promotion of open and scalable technologies continues to be the focus at AMD,” said Raja Koduri, chief architect in the Radeon Technologies Group at AMD.
“As a member of the Khronos Group, AMD is proud to collaborate with hardware and software industry leaders to develop the Vulkan API to ignite the next evolution in PC game development.”
Having big players in the graphics world working together could mean more powerful and more efficient GPU hardware and software. µ
Thanks to a hard-coded Nvidia Tegra X1 flaw
Time's up. Me too. Not him
Redmond says 'the fix is more complex than initially anticipated'
And, yep, they're really expensive