Technical Papers Archive

SpaDA: A Spatial Dataflow Architecture Programming Language


Authors: Lukas Gianinazzi (Noéda), Tal Ben-Nun (Lawrence Livermore National Laboratory (LLNL)), and Torsten Hoefler (ETH Zürich)

Abstract: Spatial dataflow architectures like the Cerebras Wafer-Scale Engine deliver exceptional performance in AI and scientific computing by distributing scratchpad memory across hundreds of thousands of processing elements (PEs). Yet programming these architectures remains difficult: with no shared memory, data movement requires explicit configuration, and asynchronous task management introduces substantial complexity. We present SpaDA, a programming language that offers precise control over data placement, dataflow patterns, and asynchronous operations while abstracting low-level architectural details. We design and implement a compiler targeting Cerebras CSL through multi-level lowering and unique optimization passes. SpaDA functions as a high-level programming interface and an intermediate representation for domain-specific languages (DSLs), demonstrated here with the GT4Py stencil DSL. SpaDA enables concise expression of operations with complex parallel patterns — including collective operations, multi-dimensional stencils, dense and sparse linear algebra — in 7.49× fewer lines than CSL, achieving over 260 TFlop/s across 730,000 PEs on a single device.


Presentation: file

Back to Technical Papers Archive Listing