3 citations found. Retrieving documents...
R. Pozo. Template Numerical Toolkit (TNT) for Linear Algebra. National Insitute of Standards and Technology.

 Home/Search   Document Not in Database   Summary   Related Articles   Check  

This paper is cited in the following contexts:
A Modern Framework for Portable High Performance Numerical Linear.. - Siek (1999)   (6 citations)  (Correct)

....algebra library, Blitz [58] is related to MTL in that it uses template metaprogramming [55] techniques to achieve high performance. In addition, the use of expression templates was pioneered in Blitz . CHAPTER 3. RELATED WORK 16 Library OO ET Op Generative Template Numeric Toolkit (TNT) [47] X X C Scientific Library (SL ) 20] X X Generative Matrix Computation Library (GMCL) 16] X X X LAPACK [25] X Sparslib [26] X GNU Scientific Software Library (GNUSSL) 46] X Newmat [19] X X Table 3.1. Summary of C Libraries for Linear Algebra 3.4 Generic Programming and Software ....

....unique among linear algebra libraries because each algorithm (for the most part) is implemented with just one template function. From a software maintenance standpoint, the reuse of code gives MTL a significant advantage over the BLAS [22, 23, 36] or even other object oriented libraries like TNT [47] (which still has different subroutines for different matrix formats) Because of the code reuse provided by generic programming, MTL has an order of magnitude fewer lines of code than the Netlib Fortran BLAS [1] while providing much greater functionality and achieving significantly better ....

[Article contains additional citation context not shown here]

R. Pozo. Template Numerical Toolkit (TNT) for Linear Algebra. National Insitute of Standards and Technology.


The Matrix Template Library: A Unifying Framework for.. - Siek, Lumsdaine (1998)   (3 citations)  (Correct)

....with just one template function. A single algorithm is used whether the matrix is sparse, dense, banded, is single precision, double, complex, etc. From a software maintenance standpoint, this reuse of code gives MTL a huge advantage over the BLAS or even other object oriented libraries like TNT [10] (which has different algorithms for different matrix formats) The generic algorithm code reuse results in the MTL having a 6.5 times fewer lines of code than the netlib Fortran BLAS. The MTL has 8,399 lines of code for the algorithms and 15,411 lines of code for components, for a total of 23,810 ....

....vector product. The compilers used were Kuck and Associates C [5] and the Sun Solaris C compiler with maximum available optimizations. The experiments were run on a Sun UltraSPARC 170E. Fig. 4 shows the dense matrix vector performance for MTL, Fortran BLAS, the Sun Performance Library, and TNT [10]. 0 50 100 150 200 250 300 350 400 450 0 50 100 150 N MTL Fortran BLAS Sun Perf Lib TNT Fig. 4. Dense Matrix Vector Multiply (Column Oriented) Fig. 5 shows the sparse matrix vector performance for MTL, SPARSKIT [12] Fortran) NIST [11] C) and TNT (C ) The sparse matrices used are from the ....

[Article contains additional citation context not shown here]

Roldan Pozo. Template Numerical Toolkit (TNT) for Linear Algebra. National Insitute of Standards and Technology.


Generic Programming for High Performance Numerical Linear.. - Siek, Lumsdaine, Lee (1998)   (9 citations)  (Correct)

....the matrix is sparse, dense, banded, single precision, double, complex, etc. From a software maintenance standpoint, the reuse of code gives MTL a significant advantage over the Generic Programming for Numerical Linear Algebra 4 BLAS [6, 7, 13] or even other object oriented libraries like TNT [16] (which has different algorithms for different matrix formats) Because of the code reuse provided by generic programming, MTL has an order of magnitude fewer lines of code than the Netlib Fortran BLAS [20] while providing much greater functionality and achieving significantly better ....

....in [19] 6 Performance Experiments In the following figures we present some performance results comparing MTL with other available libraries (both public domain and vendor supplied) Fig. 4 shows the dense matrix matrix product performance for MTL, Fortran BLAS, the Sun Performance Library, TNT [16], and ATLAS [17] all obtained on a Sun UltraSPARC 170E. The MTL and TNT executables were compiled using Kuck and Associates C (KCC) 11] in conjunction with the Solaris C compiler. ATLAS was compiled with the Solaris C compiler and the Fortran BLAS (obtained from Netlib) were compiled with the ....

Roldan Pozo. Template Numerical Toolkit (TNT) for Linear Algebra. National Insitute of Standards and Technology.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC