Expand description
Abstracts data-parallel computation based on a map operation for a given job.
Map module enables efficient execution of the same device kernel across
multiple chunks of input data in parallel. It is designed to work within a job
context, leveraging hardware-level parallelism in XCENA device.
It applies the same device kernel across each chunk of the input data, leveraging hardware-level parallelism for performance.
This module provides a Rust abstraction around the C++ pxl::runtime::Map class.
Structsยง
- Map
- Abstracts data-parallel computation based on a map operation for a given job.