An overview of the sparse basic linear algebra subprograms: The new standard from the BLAS technical forum

By Iain S. Duff, M. Heroux, Roldan Pozo. 2002

177 citations recorded by the source corpus.

Subfields: Computational Theory and Mathematics, Computer Networks and Communications, Hardware and Architecture

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Computer Networks and Communications

In Hardware and Architecture

Citations in the corpus

Optimal combination forecasts for hierarchical time series (2011, methodology): “(1968) and Edwards and Orcutt (1969) argued that information loss is substantial in aggregation and therefore the bottom-up method gives more accurate forecasts. Shlifer and Wolff (1979) compared the forecasting performance of both methods and concluded that the bottom-up method is preferable under ”
OSKI: A library of automatically tuned sparse matrix kernels (2005, methodology): “Related Work: Approaches that Complement Libraries The Sparse BLAS Standard [13] inspired the OSKI design.”
A thermal energy storage process for large scale electric applications (2010, methodology): “[16] Iain S. Duff, Michael A. Heroux, and Roldan Pozo, “An overview of the Sparse Basic Linear Algebra Subprograms: The new standard from the BLAS Technical Forum,” ACM TRANS.”
Model Pruning Enables Efficient Federated Learning on Edge Devices (2019, unknown): “using Sparse BLAS [5].”
Automatic performance tuning of sparse matrix kernels (2003, methodology): “The interested reader can find discussion of these topics in presentations by Saad [268], in the “Templates” guide to solving linear systems [30], and in Duff’s book on sparse LU factorization [108], among other sources [109, 267, 254].”
The Combinatorial BLAS: design, implementation, and applications (2011, methodology): “The Sparse BLAS (Duff et al., 2002) is a standard API for numerical matrix- and vector-level primitives; its focus is infrastructure for iterative linear system solvers, and therefore it does not include such primitives as sparse matrix– matrix multiplication and sparse matrix indexing.”
OuterSPACE: An Outer Product Based Sparse Matrix Multiplication Accelerator (2018, background): “It is the general consensus that a selection of linear algebra routines optimized over years of research can be used to accelerate a wide range of graph and scientific algorithms [22].”
Cambricon-X: An accelerator for sparse neural networks (2016, background): “Compared against the CPU with the sparse library (i.e., Sparse BLAS [27]), on average, our accelerator achieves 144.41x speedup.”
Computing the Action of the Matrix Exponential, with an Application to Exponential Integrators (2011, background): “The advantage of working with B is the ability to invoke level 3 BLAS [6] [7], which should lead to faster execution.”