Styles for this page:[dark on bright background][bright on dark background]

Michele Martone's (transitory) home page

I am a Ph.D. graduate in Computer Engineering at Tor Vergata University of Rome, Italy.

Currently, I am part of the High Level Support Team at the Max Planck Institute for Plasma Physics in Garching, Germany ( homepage there ).

Recently (Autumn'09-Spring'10), I've been working as a visiting student in Poland at the IBS PAN institute, under tutorship of Marcin Paprzycki.

My advisor was prof. Salvatore Tucci.

My work focuses on High Performance and Scientific Computing (namely, in the field of sparse matrices computations).

I am an eager GNU/Linux user and free software enthusiast and I am interested in the ways to make computer code, (especially scientific!) run fast and efficiently.

I have an Erdös number (5), thanks to Marcin Paprzycki and Salvatore Filippone.

Contacts

e-Mail address (new) : michele.martone_strip_off_this_junk_please@ipp_this_too_.mpg.de .

e-Mail address (old) : michele.martone_strip_off_this_junk_please@uniroma2_this_too_.it .

Publications

I also have a record on DBLP.
[7] Michele Martone, Marcin Paprzycki, and Salvatore Filippone. An improved sparse matrix-vector multiply based on recursive sparse blocks layout. In I. Lirkov, S. Margenov, and J. Waśniewski, editors, LSSC 2011, LNCS 7116, pages 606-613. Springer, Heidelberg, 2012.
[6] Michele Martone, Salvatore Filippone, Marcin Paprzycki, and Salvatore Tucci. About the assembly of recursive sparse matrices. In Proceedings of the International Multiconference on Computer Science and Information Technology, Wisla. Poland, October 2010.
Recently, we have introduced an approach to multi-core computations on sparse matrices using recursive partitioning, called Recursive Sparse Blocks (RSB). In this document, we discuss issues involved in assembling matrices in the RSB format. Since the main expected application area is iterative methods, we compare the performance of matrix assembly to that of matrix-vector multiply (SPMV), outlining both scalability of the method and execution times ratio.
[5] Michele Martone, Salvatore Filippone, Pawel Gepner, Marcin Paprzycki, and Salvatore Tucci. Use of hybrid recursive CSR/COO data structures in sparse matrices-vector multiplication. In Proceedings of the International Multiconference on Computer Science and Information Technology, Wisla, Poland, October 2010.
Recently, we have introduced an approach to basic sparse matrix computations on multicore cache based machines using recursive partitioning. Here, the memory representation of a sparse matrix consists of a set of submatrices, which are used as leaves of a quad-tree structure. In this paper, we evaluate the performance impact, on the Sparse Matrix-Vector Multiplication (SPMV), of a modification to our Recursive CSR implementation, allowing the use of multiple data structures in leaf matrices (CSR/COO, with either 16/32 bit indices).
[4] Michele Martone, Salvatore Filippone, Marcin Paprzycki, and Salvatore Tucci. On BLAS operations with recursively stored sparse matrices. In Proceedings of the International Symposium on Symbolic and Numeric Algorithms for Scientific Computing, Timisoara, Romania, September 2010.
Recently, we have proposed a recursive partitioning based layout formulti-core computations on sparse matrices. Based on positive results of our initial experiments with matrix-vector multiplication, we discuss how this storage format can be utilized across a range of BLAS-style matrix operations.
[3] Michele Martone, Salvatore Filippone, Marcin Paprzycki, and Salvatore Tucci. On the usage of 16 bit indices in recursively stored sparse matrices. In Proceedings of the International Symposium on Symbolic and Numeric Algorithms for Scientific Computing, Timisoara, Romania, September 2010.
In our earlier work, we have investigated on the feasibility of utilization of recursive partitioning in sparse matrix computations on multi-core cache based computers. After having assessed its effectiveness in matrix-vector multiplication in [4], here we discuss how the use of this storage format can be tuned by limiting the bandwidth overhead by the usage of 16 bit indices in leaf submatrices (at the end of the recursion). Experimental results obtained on a collection of matrices and on 8-core machines illustrate effectiveness of the proposed approach.
[2] Michele Martone, Salvatore Filippone, Salvatore Tucci, Marcin Paprzycki, and Maria Ganzha. Utilizing recursive storage in sparse matrix-vector multiplication - preliminary considerations. In Philips [1], pages 300-305.
Computations with sparse matrices on multicore cache-based computers are affected by the irregularity of the problem at hand, and performance degrades easily. In this note we propose a recursive storage format for sparse matrices, and evaluate its usage for the Sparse Matrix-Vector (SpMV) operation on two multicore and one multiprocessor machines. We report benchmark results showing high performance and scalability comparable to current state of the art implementations.
[1] Thomas Philips, editor. Proceedings of the ISCA 25th International Conference on Computers and Their Applications, CATA 2010, March 24-26, 2010, Sheraton Waikiki Hotel, Honolulu, Hawaii, USA. ISCA, 2010.
Citations available in BibTeX format.

Theses

Software (academic)

Software (personal)

GNU/Linux Software Configs

Misc stuff

Public GPG key

Here is my GPG public key. You can import it in your keyring issuing:

gpg --search 'Michele Martone'

or

wget -O - http://claudius.ce.uniroma2.it/~martone/michele.martone.asc | gpg --import

in your shell.

Personal


Viewable With Any Browser

Valid HTML 4.0 Transitional