VK_EXT_gpu_assisted_validation overhead and default activation state

The VK_EXT_gpu_assisted_validation extension allows the Vulkan SDK to offload validation checks to the GPU, which reduces the CPU bottleneck typically associated with heavy validation layers and improves the accessibility of the debugging workflow.

While this feature provides immediate feedback during rendering, it introduces constraints regarding GPU memory consumption and potential performance degradation on low-end hardware. Currently, the extension is opt-in and requires explicit activation via VK_VALIDATION_FEATURE_ENABLE_GPU_ASSISTED_EXT during instance creation.

There is an unresolved design decision regarding the default state of this validation mode across different build configurations. Specifically, it is unclear if this should remain a manual opt-in or become a default for debug-specific builds.

• What are the criteria for determining if GPU-assisted validation should be enabled by default in debug environments?
• How can the memory overhead be mitigated for devices with limited VRAM when this extension is active?