XTC, A Research Platform for Optimizing AI Workload Operators
By: Pompougnac Hugo , Guillon Christophe , Noiry Sylvain and more
Achieving high efficiency on AI operators demands precise control over computation and data movement. However, existing scheduling languages are locked into specific compiler ecosystems, preventing fair comparison, reuse, and evaluation across frameworks. No unified interface currently decouples scheduling specification from code generation and measurement. We introduce XTC, a platform that unifies scheduling and performance evaluation across compilers. With its common API and reproducible measurement framework, XTC enables portable experimentation and accelerates research on optimization strategies.
Similar Papers
Agentic Operator Generation for ML ASICs
Distributed, Parallel, and Cluster Computing
Creates code for new computer chips automatically.
Kant: An Efficient Unified Scheduling System for Large-Scale AI Clusters
Distributed, Parallel, and Cluster Computing
Makes AI computers work faster and smarter.
Holistic Heterogeneous Scheduling for Autonomous Applications using Fine-grained, Multi-XPU Abstraction
Operating Systems
Makes self-driving cars faster and safer.