Case Study: Barefoot Networks achieves AI-driven, high-performance programmable networking with Franz Inc.'s Allegro CL

Barefoot Networks Uses Franz's Allegro CL to Drive Artificial Intelligence in Networks

Barefoot Networks turned to Franz Inc.'s Allegro CL® (Allegro Common Lisp) to bring artificial intelligence into its Tofino series of programmable switch chips. The challenge was to manage extreme design complexity—creating embedded languages for netlist generation, gate placement and routing—and to scale custom module generators and layout compilers to achieve better chip area, performance and power efficiency while enabling ML-driven network analytics.

Franz Inc.'s Allegro CL provided a powerful Common Lisp environment for symbolic manipulation and homoiconicity, allowing Barefoot to build the embedded languages and large-scale compilers needed for Tofino. The result: Tofino switches can run in‑band, line‑rate machine learning that analyzes every packet and detects anomalies in real time with nanosecond accuracy, yielding measurable competitive advantages in chip area, performance and power efficiency.

Barefoot Networks

Patrick Bosshart

CTO and Co-founder