Jacket DLA

Double Precision Linear Algebra

Jacket DLA provides dense linear algebra support for the base Jacket product. Functions for matrix analysis, linear systems solutions, eigen analysis, singular value decomposition, and various factorizations are enabled for both single- and double-precision as well as for complex values with Jacket DLA. AccelerEyes, our partners, and the CUDA community of developers are continuing to expand support for high performance linear algebra in every release so check our documentation and release notes for the latest functions supported.

The following MATLAB functions are currently enabled with JacketDLA:

■ MLDIVIDE - Least-squares solutions to over- and under-determined systems

■ LU - LU decomposition

■ SVD - Singular value decomposition

■ EIG - Eigenvalues

■ QR - QR factorization

■ CHOL - Cholesky decomposition

■ NV - Matrix inverse

■ MPOWER - Matrix power

■ NORM - L2 matrix norm

■ DET - Matrix determinant

■ HESS - Hessenberg form of matrix

For applications that do not require double-precision, the base Jacket license provides limited, single precision versions of matrix power, SVD, LU, QR, and MLDIVIDE.

雙精度線性代數

Jacket DLA提供基礎Jacket產品密集的線性代數支持。矩陣分析，線性系統的解決方案，特徵分析，奇異值分解，以及各種因子分解的功能單一精度和雙精度都可以啟用以及Jacket DLA複數。AccelerEyes，我們的合作夥伴，開發者的CUDA社會正在在每一個版本的高性能線性代數繼續擴大它的支持，请观看我們的文檔和發行說明寻找最新支持的功能。

一下的MATLAB功能目前與 JacketDLA啟用：

■ MLDIVIDE - Least-squares solutions to over- and under-determined systems

■ LU - LU decomposition

■ SVD - Singular value decomposition

■ EIG - Eigenvalues

■ QR - QR factorization

■ CHOL - Cholesky decomposition

■ NV - Matrix inverse

■ MPOWER - Matrix power

■ NORM - L2 matrix norm

■ DET - Matrix determinant

■ HESS - Hessenberg form of matrix

對於不需要雙精度的應用，基礎Jacket許可證提供有限單一精度矩陣能力，SVD，LU，QR，並 MLDIVIDE。