AMD's GPUOpen Posts New Vulkan Memory Allocator
AMD's GPUOpen initiative has posted a number of Vulkan open-source projects over time from the Anvil Vulkan framework to a Vulkan-supported CodeXL and various code samples. Their latest open-source project is a Vulkan memory allocator.
Details are still light as GPUOpen doesn't appear to have officially announced it yet, I just stumbled upon it through my routine Vulkan crawling around GitHub. The initial public commit to GPUOpen's VulkanMemoryAllocator was back on Friday.
The VulkanMemoryAllocator is designed as a "easy to integrate Vulkan memory allocation library." This allocation library tries to remove some of the boilerplate code, remove additional indirection, make it easier to manage memory across drivers, out-of-memory handling, and more. The library is cross-platform and is a self-contained C++ library.
Right now the documentation is fairly light with the README just encouraging interested developers to read the code comments in this header file.
Details are still light as GPUOpen doesn't appear to have officially announced it yet, I just stumbled upon it through my routine Vulkan crawling around GitHub. The initial public commit to GPUOpen's VulkanMemoryAllocator was back on Friday.
The VulkanMemoryAllocator is designed as a "easy to integrate Vulkan memory allocation library." This allocation library tries to remove some of the boilerplate code, remove additional indirection, make it easier to manage memory across drivers, out-of-memory handling, and more. The library is cross-platform and is a self-contained C++ library.
Right now the documentation is fairly light with the README just encouraging interested developers to read the code comments in this header file.
5 Comments