I feel like an API like that would be nice to have and citing MDN, creating pipelines asynchronously "is generally preferable"

We currently create all pipelines and resources lazily synchronously but maybe the future is asynchronous? I feel like that could enable us to distribute the work nicely (for example when creating multiple pipelines one can begin asynchronously being created by WebGPU API while we begin resolving the next)
I feel like an API like that would be nice to have and citing MDN, creating pipelines asynchronously "is generally preferable"

We currently create all pipelines and resources lazily synchronously but maybe the future is asynchronous? I feel like that could enable us to distribute the work nicely (for example when creating multiple pipelines one can begin asynchronously being created by WebGPU API while we begin resolving the next)