MLCommons Chakra: Advancing Performance Benchmarking and Co-design using Standardized Execution Traces
For AI/ML researchers and system designers, Chakra provides a standardized representation to enable reproducible performance analysis and software-hardware co-design, addressing the need for agile benchmarking in rapidly evolving AI systems.
Chakra introduces an open, portable ecosystem for performance benchmarking and co-design of distributed ML workloads, using standardized execution traces (ETs) that represent compute, memory, and communication operations. The ecosystem includes tools for collection, analysis, generation, and adoption of ETs, with adoption by MLCommons and industry partners.
The fast pace of artificial intelligence~(AI) innovation demands an agile methodology for observation, reproduction and optimization of distributed machine learning~(ML) workload behavior in production AI systems and enables efficient software-hardware~(SW-HW) co-design for future systems. We present Chakra, an open and portable ecosystem for performance benchmarking and co-design. The core component of Chakra is an open and interoperable graph-based representation of distributed AI/ML workloads, called Chakra execution trace~(ET). These ETs represent key operations, such as compute, memory, and communication, data and control dependencies, timing, and resource constraints. Additionally, Chakra includes a complementary set of tools and capabilities to enable the collection, analysis, generation, and adoption of Chakra ETs by a broad range of simulators, emulators, and replay tools. We present analysis of Chakra ETs collected on production AI clusters and demonstrate value via real-world case studies. Chakra has been adopted by MLCommons and has active contributions and engagement across the industry, including but not limited to NVIDIA, AMD, Meta, Keysight, HPE, and Scala, to name a few.