GPUOpen's Vulkan Memory Allocator 2.1 Being Prepped With Many Additions
AMD's GPUOpen group has released their first beta of the Vulkan Memory Allocator 2.1 release after "many months of development" and as such comes with many new features.
VulkanMemoryAllocator as a refresher is the open-source AMD effort to provide an easy-to-use and integrate Vulkan memory allocation library to ease the process of bringing up new Vulkan code. The VulkanMemoryAllocator is used by the likes of Google's Filament renderer, vkDoom3, LWJGL, the Anvil framework, and others.
While retaining API compatibility with the current VulkanMemoryAllocator. the 2.1 beta release adds support for a linear allocation algorithm, support to record a sequence of calls to the allocator and to be able to replay that later, better support for non-coherent memory, improved debugging, custom pool changes, and a variety of other additions.
More details on the 2.1 beta of the VulkanMemoryAllocator can be found via GitHub.
VulkanMemoryAllocator as a refresher is the open-source AMD effort to provide an easy-to-use and integrate Vulkan memory allocation library to ease the process of bringing up new Vulkan code. The VulkanMemoryAllocator is used by the likes of Google's Filament renderer, vkDoom3, LWJGL, the Anvil framework, and others.
While retaining API compatibility with the current VulkanMemoryAllocator. the 2.1 beta release adds support for a linear allocation algorithm, support to record a sequence of calls to the allocator and to be able to replay that later, better support for non-coherent memory, improved debugging, custom pool changes, and a variety of other additions.
More details on the 2.1 beta of the VulkanMemoryAllocator can be found via GitHub.