P4 Developer Days – NSDI 2025 Alkali

Register to attend this P4 Developer Days webinar, “NSDI 2025 Alkali”

Date: October 1, 2025
Time: 9:00am Pacific

Abstract

Programming SmartNICs today is notoriously difficult. NICs from different vendors—or even different generations of the same vendor—exhibit significant variation in hardware parallelism, memory hierarchies, and interconnects. As a result, porting programs across NICs is labor-intensive and requires extensive manual refactoring to meet performance expectations on each target.

In this presentation, we will present the demo and design of Alkali, a SmartNIC compilation framework that enables developers to write target-independent programs, while the compiler automatically handles cross-NIC porting and performance tuning. Alkali achieves this by: (1) introducing a novel intermediate representation (IR) that supports building a reusable and extensible compiler across diverse NICs, and (2) developing an optimization algorithm that automatically transforms and parallelizes programs based on the target NIC’s hardware characteristics.

Alkali is built on the MLIR infrastructure and is fully open source (https://github.com/utnslab/Alkali). It currently supports four distinct NIC backends and continues to grow, including future support for P4 as part of an effort to make SmartNIC programming more portable and productive.

Speakers

Jiaxin Lin is an incoming assistant professor at Cornell University. She received her Ph.D. from UT Austin in the summer of 2025. Her research aims to design innovative software and hardware for accelerated data center networks. She has received the Google and Meta PhD Fellowships in Computer Networking and was selected as an MIT EECS Rising Star in 2024.

 

 

Zhiyuan Guo is a final-year PhD student at UC San Diego. His research focuses on co-designing datacenter systems and applications for resource disaggregation. He is actively extending the concept of disaggregation and cohesive design to various datacenter systems, including remote memory, networking accelerators, and machine learning infrastructure.

 

Register to attend this webinar!

Share the Post:

Contact us

Fill out the form below, and we will be in touch shortly.