BlueField DOCA Example ( public ) ( TCP and P4 Programmable Data Plane Switches )
This lab explores the fundamentals of invoking hardware accelerators using the DOCA Core library and the DOCA Execution Model on the NVIDIA BlueField DPU. The lab exercise focuses on configuring and utilizing the DOCA Compression Accelerator through the doca_compress API. The objective is to demonstrate how to create and submit compression tasks asynchronously and implement completion and error callbacks to handle task results. The lab provides hands-on experience with accelerator invocation via the DOCA Execution Model. A working knowledge of C programming is a recommended prerequisite.
4
0
(0)
0
Oct. 13, 2025, 8:52 p.m.
Oct. 13, 2025, 8:52 p.m.
Authors
- Amith Gorthi Srinivasa Prabhakara Narasimha , University of South Carolina (Amithgspn@sc.edu)
- Elie Kfoury , University of South Carolina (ekfoury@email.sc.edu)