# DISTAL: The Distributed Tensor Algebra Compiler

arXiv: https://arxiv.org/abs/2203.08069 · HTML: https://mlsanity.com/paper/2203.08069/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Mar 2022.

## Built on by (most cited first, 6 of 6)

- [WaferLLM: Large Language Model Inference at Wafer Scale](https://mlsanity.com/paper/2502.04563/) (2025)
- [Improving Parallel Program Performance with LLM Optimizers via Agent-System Interfaces](https://mlsanity.com/paper/2410.15625/) (2024)
- [PartIR: Composing SPMD Partitioning Strategies for Machine Learning](https://mlsanity.com/paper/2401.11202/) (2024)
- [Scaling Deep Learning Computation over the Inter-Core Connected Intelligence Processor with T10](https://mlsanity.com/paper/2408.04808/) (2024)
- [Serving Deep Learning Model in Relational Databases](https://mlsanity.com/paper/2310.04696/) (2023)
- [Masked Matrix Multiplication for Emergent Sparsity](https://mlsanity.com/paper/2402.14118/) (2024)

## Abstract (arXiv)

We introduce DISTAL, a compiler for dense tensor algebra that targets modern distributed and heterogeneous systems. DISTAL lets users independently describe how tensors and computation map onto target machines through separate format and scheduling languages. The combination of choices for data and computation distribution creates a large design space that includes many algorithms from both the past (e.g., Cannon's algorithm) and the present (e.g., COSMA). DISTAL compiles a tensor algebra domain specific language to a distributed task-based runtime system and supports nodes with multi-core CPUs and multiple GPUs. Code generated by DISTAL is competitive with optimized codes for matrix multiply on 256 nodes of the Lassen supercomputer and outperforms existing systems by between 1.8x to 3.7x (with a 45.7x outlier) on higher order tensor operations.