SPIR-V Float16 capability (Turing)

Is there any way to get Turing to consume SPIR-V shaders which use Float16?
I know the hardware should be able to do it, but there doesn’t seem to be an option to actually enable it.

(I run into “SPIR-V module not valid: Using a 16-bit floating point type requires the Float16 or Float16Buffer capability, or an extension that explicitly enables 16-bit floating point.”)

To partially answer my own question: I believe the way to do this correctly would be to use the VkPhysicalDeviceShaderFloat16Int8FeaturesKHR extension (from Vulkan 1.1.115) to ask for shaderFloat16 capabilities.

However, there currently doesn’t seem to be a driver that actually exposes this (which is understandable given how new it is – hopefully soon!).