WRSWorld Research StateContribute
WORLD RESEARCH STATE

GSPMD: compiler propagation of tensor sharding

Contribution grade

Distributed ML compiler systems · Human-designed research method or resource

Reference score145,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Inferred operator partitioning from limited tensor annotations and generated SPMD computation, supporting uneven/nested sharding, padding and halo exchange in XLA.

Scientific scope

Inferred operator partitioning from limited tensor annotations and generated SPMD computation, supporting uneven/nested sharding, padding and halo exchange in XLA. Extends GShard and inherits XLA/SPMD. Pipeline support requires wrappers or existing pipelining in some cases; annotations are not removed entirely. Tested TPU utilization and compiler scalability are not general hardware/workload optimality.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Extends GShard and inherits XLA/SPMD. Pipeline support requires wrappers or existing pipelining in some cases; annotations are not removed entirely. Tested TPU utilization and compiler scalability are not general hardware/workload optimality.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction