topic Significant performance improvements in Intel® oneMKL PARDISO 2026.0 in Intel® oneAPI Math Kernel Library https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Significant-performance-improvements-in-Intel-oneMKL-PARDISO/m-p/1755532#M37641 <P>Intel® oneMKL PARDISO 2026.0 delivers significant performance improvements for direct sparse solvers.</P> <P><BR />Key improvements:</P> <P>- Reordering (Phase1): Up to 2.7x faster with improved paralllel nested dissection<BR />- Geometric mean speedup: 1.7x across diverse matrices<BR />- Parallel nested dissection now the default (iparm[1]=3)<BR />- CNR mode now supports parallel version (no performance penalty for reproducible results)</P> <P>- Factorization (Phase 2): Up to 12x faster<BR />- Geometric mean speedup: 2.5x<BR />- New optimization option (iparm[1]=43) provides up to 2.3x additional speedup for large matrices.</P> <P>&nbsp;</P> <P>For more details, see this article:</P> <P><A href="https://www.intel.com/content/www/us/en/developer/articles/technical/pardiso-2026-performance-breakthrough.html" target="_blank">https://www.intel.com/content/www/us/en/developer/articles/technical/pardiso-2026-performance-breakthrough.html</A></P> Tue, 04 Aug 2026 00:39:54 GMT Chao_Y_Intel 2026-08-04T00:39:54Z Significant performance improvements in Intel® oneMKL PARDISO 2026.0 https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Significant-performance-improvements-in-Intel-oneMKL-PARDISO/m-p/1755532#M37641 <P>Intel® oneMKL PARDISO 2026.0 delivers significant performance improvements for direct sparse solvers.</P> <P><BR />Key improvements:</P> <P>- Reordering (Phase1): Up to 2.7x faster with improved paralllel nested dissection<BR />- Geometric mean speedup: 1.7x across diverse matrices<BR />- Parallel nested dissection now the default (iparm[1]=3)<BR />- CNR mode now supports parallel version (no performance penalty for reproducible results)</P> <P>- Factorization (Phase 2): Up to 12x faster<BR />- Geometric mean speedup: 2.5x<BR />- New optimization option (iparm[1]=43) provides up to 2.3x additional speedup for large matrices.</P> <P>&nbsp;</P> <P>For more details, see this article:</P> <P><A href="https://www.intel.com/content/www/us/en/developer/articles/technical/pardiso-2026-performance-breakthrough.html" target="_blank">https://www.intel.com/content/www/us/en/developer/articles/technical/pardiso-2026-performance-breakthrough.html</A></P> Tue, 04 Aug 2026 00:39:54 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/Significant-performance-improvements-in-Intel-oneMKL-PARDISO/m-p/1755532#M37641 Chao_Y_Intel 2026-08-04T00:39:54Z