Software Engineer TT-Fabric
Posted Updated
Description
You will architect, implement, and maintain a low-level networking library for distributed inference and training. You will design scalable communication systems, optimize protocols and synchronization, integrate APIs into the programming model, and help define the distributed systems architecture.
Responsibilities
- Architect, implement, and maintain the TT-Fabric low-level networking library
- Design scalable communication systems for thousands of AI processors
- Optimize protocols, synchronization strategies, and data movement
- Integrate TT-Fabric APIs into the broader programming model
- Help define the long-term distributed systems architecture
Requirements
- Deep C or C++ experience
- Low-level or bare-metal systems engineering experience
- Knowledge of hardware-software interaction
- Performance tuning expertise
- Knowledge of networking, synchronization, and large-cluster communication
