20 results for “Understanding of automatic parallelism systems”
CS papers onlyHybrid search: Keyword + semantic, ranked by combined score.ⓘ
Want pure semantic search? Try claim verification →
This paper compares the performance of four parallelization variants of the tiled Cholesky decomposition using OpenMP and HPX runtimes.
This paper introduces the Scatter-Gather parallel-programming and parallel-execution model in the form of a simple imperative language.
This paper proposes a new family of models for in-place parallel and distributed computation that ensures efficiency and in-place guarantees at all processor counts.
This paper presents Wavelet, the first formally verified compiler for asynchronous dataflow architectures, using a capability type system and Lean formalization.
The paper implements and evaluates several hardware priority queue architectures on modern FPGA platforms and provides a quantitative analysis.
This paper surveys the evolution of massively parallel systems over the past four decades, focusing on hardware architectures, interconnection networks, and synchronization primitives.
This paper presents a simulation framework to optimize workflow composition in high-throughput computing environments, demonstrating up to 3.8x throughput increase and a 14.9x reduction in network ove…
Elam Cohavi, Nicolas Bohm Agostini, Jude Haris, Antonino Tumeo +2 more
This paper optimizes data transfer between host and accelerator in machine learning workloads by eliminating redundant staging operations.
This paper analyzes data access patterns in large language model workloads on multi-partition GPUs and presents optimization strategies based on inter-workgroup sharing patterns.
This paper presents a practical MPI malleability methodology for HPC clusters.
Zepeng Li, Jie Ren, Zhanyong Tang, Jie Zheng +1 more
AutoPass is a multi-agent framework for compiler performance tuning using a large language model, enabling it to query compiler-internal optimization states and analyze the intermediate representation…
Zhiqiang Yuan, Chong Wang, Zhenhao Zhou, Junwei Liu +1 more
This paper studies the performance gap between C and Rust code generated by a transpiler and proposes an LLM-based approach to recover the performance of translated Rust.
Jiasheng Zhou, Longbin Zeng, Clavis Chen, Ruiming Lu +4 more
ARGUS is a low-overhead, fine-grained, always-on tracing and real-time analysis system for training workloads in large production clusters.