Sparse 2025
who it's for
sparse ML researchers, compiler engineers, GPU kernel developers, high-performance computing engineers, quantum simulation researchers, distributed systems engineers, formal verification researchers
topics
structured sparsity compilation; polyhedral indexing for sparsity; sparsity-aware autoscheduling; Fast Compressed Sparse Row; sparse tensor GPU kernels; indirect einsum lowering; sparse quantum simulation; distributed sparse SciPy; tile fusion for sparse matmul; matrix reassociation optimizations; relational separation logic for arrays
| title | author | length |
|---|---|---|
| Compiling and Compressing Structured Tensors | Emilien Bauer | 15m |
Two systems compile symbolic sparsity into executable code: an earlier DSL-driven compiler called Struct that infers structure inference (diagonals, symmetry… | ||
| Optimizations and abstractions for sparse machine learning | Charith Mendis | 19m |
Holistic optimizations trade between sparse and dense work using matrix reassociations and an offline/online compiler called sensei that picks the fastest… | ||
| Sparsity-Aware Autoscheduling for Numpy with Finch and Galley | Willow Ahrens | 19m |
Galley implements an autoscheduler for the Finch sparse tensor compiler that makes concrete decisions: it uses database-style cardinality estimation (notably… | ||
| Quantum Simulation with Sparse Tensors | Meisam Tarabkhah | 20m |
A compiler framework called Lightweight Functional Quantum Simulator (LFQS) converts gate-level linear algebra into an intermediate language (SDQL/SQL) and… | ||
| Sparse Computing Drives Energy-Efficient Computation for Artificial Intelligence | Guohao Dai | 23m |
Presents a GPU-focused sparse-kernel library dgSPARSE with an auto-selection model that outperforms commercial kernels across GNN models and sparse operators… | ||
| Hyperreal Specifications for Continuous Sparse Data Computations | Vladimir Gladshtein | 16m |
A new relational separation logic called LGTM∞ specifies array-encoded geometric data as code (getter procedures) and generalises hypersafety to real-indexed… | ||
| Distributed Sparse Computing with Legate Sparse | Rohan Yadav | 17m |
Legate Sparse transparently distributes unmodified SciPy sparse programs across CPU/GPU clusters by composing with the distributed NumPy library cuPyNumeric… | ||
| Loop Fusion in Matrix Multiplications with Sparse Dependence | Kazem Cheshmi | 18m |
A compiler pass called tile fusion fuses tiles of two consecutive matrix multiplications when sparsity creates partial dependences, implemented as a… | ||
| Insum: Sparse GPU Kernels Simplified and Optimized with Indirect Einsums | Saman Amarasinghe | 15m |
Introduces Insum, a compiler that encodes sparse formats as indirect Einsums and lowers them into an extended PyTorch/Torch-Compiler pipeline to generate… | ||
| nothing matches | ||