O
Oded Schwartz
Researcher at Hebrew University of Jerusalem
Publications - 75
Citations - 2774
Oded Schwartz is an academic researcher from Hebrew University of Jerusalem. The author has contributed to research in topics: Matrix multiplication & Strassen algorithm. The author has an hindex of 28, co-authored 73 publications receiving 2540 citations. Previous affiliations of Oded Schwartz include Technical University of Berlin & Tel Aviv University.
Papers
More filters
Journal ArticleDOI
Minimizing Communication in Linear Algebra
TL;DR: This work generalizes a lower bound on the amount of communication needed to perform dense, n-by-n matrix multiplication using the conventional O(n3) algorithm to a much wider variety of algorithms, including LU factorization, Cholesky factors, LDLT factors, QR factors, the Gram–Schmidt algorithm, and algorithms for eigenvalues and singular values.
Journal ArticleDOI
Minimizing Communication in Numerical Linear Algebra
TL;DR: Hong and Kung as discussed by the authors gave a lower bound on the communication complexity of matrix multiplication in the parallel case. But this lower bound was later extended to a much wider variety of linear algebra algorithms, including LU factorization, Cholesky factorization and LDLT factorization.
Journal ArticleDOI
On the complexity of approximating k -set packing
TL;DR: It is proved that the Maximumk -Set Packing problem cannot be efficiently approximated to within a factor ofmega unless P = NP, which improves the previous hardness of approximation factor of k/2^{{O({\sqrt {\ln k} })}} by Trevisan.
Journal ArticleDOI
Communication lower bounds and optimal algorithms for numerical linear algebra
TL;DR: This paper describes lower bounds on communication in linear algebra, and presents lower bounds for Strassen-like algorithms, and for iterative methods, in particular Krylov subspace methods applied to sparse matrices.
Proceedings ArticleDOI
Communication-Optimal Parallel Recursive Rectangular Matrix Multiplication
James Demmel,David Eliahu,Armando Fox,Shoaib Kamil,Benjamin Lipshitz,Oded Schwartz,Omer Spillinger +6 more
TL;DR: This work obtains the first communication-optimal algorithm for all dimensions of rectangular matrices by combining the dimension-splitting technique with the recursive BFS/DFS approach, and shows significant speedups over existing parallel linear algebra libraries both on a 32-core shared-memory machine and on a distributed-memory supercomputer.