Advances in Numerical Linear Algebra: Celebrating the Centenary of the Birth of James H. Wilkinson, May 29-30, 2019

This international workshop focuses on numerical linear algebra (NLA) and will bring together researchers working in NLA to discuss current developments, challenges in the light of evolving computer hardware and the changing needs of applications.

2019 is the centenary of the birth of James Hardy Wilkinson, FRS and Turing Prize winner (1970).  See our website about Wilkinson. Wilkinson developed the theory and practice of backward error analysis, particularly in the context of NLA, and produced detailed analyses of algorithms and software implementing those algorithms. He worked with Turing at NPL in the 1940s, and took over the development of the Pilot ACE computer when Turing left NPL. Wilkinson’s legacy endures and one of the aims of the workshop is to take stock of what has changed in error analysis since Wilkinson’s era and to discuss current challenges and ways to approach them.

Invited Speakers

Highlights of the Meeting

For a summary of the meeting and photos from it, see Highlights of Advances in Numerical Linear Algebra Conference.

Schedule

Wednesday May 29
Session 1. Chair: Nick Higham
 08.30-09.00 Registration & Coffee
 09.00-09.15 Nick Higham (The University of Manchester):
Welcome. Video (YouTube)
 09.15-09.40 Sven Hammarling (The University of Manchester):
James Hardy Wilkinson. Video (YouTube)
 09.45-10.10 Cleve Moler (MathWorks):
Eigenvalue Dating Service. Video (YouTube)
 10.15-10.40 Jack Dongarra (University of Tennessee, Oak Ridge Laboratory and University of Manchester):
Software for Linear Algebra Targeting Exascale (SLATE). Video (Yotube)
 10.45-11.15 Coffee and posters
 11.15-11.40 Volker Mehrmann (Technische Universität Berlin):
Eigenvalue computation for structured problems. A Wilkinson heritage. Video (YouTube)
 11.45-12.00 Erin Carson (Charles University):
On the Amplification of Rounding Errors. Video (YouTube)
 12.05-12.20 Philip Knight (University of Strathclyde) :
Perturbing doubly stochastic matrices. Video (YouTube)
 12.25-12.40 Srikara Pranesh (The University of Manchester):
Squeezing a Matrix Into Half Precision, with an Application to Solving Linear Systems.
Video (YouTube)
 12.45-14.00 Lunch and posters
Session 2. Chair: Jack Dongarra
14.00-14.25 Margaret Wright (Courant Institute of Mathematical Sciences, New York University):
Lessons Taught by James Wilkinson. Video (YouTube)
 14.30-14.55 Nick Trefethen (University of Oxford):
Wilkinson, Numerical Analysis, and Me. Video (YouTube)
15.00-15.25 Nick Higham (The University of Manchester):
Probabilistic Versus Worst-Case Rounding Error Analysis. Video (YouTube)
15.30-16.00 Tea and posters
 16.00-16.15 Marcus Webb (KU Leuven, Belgium):
Redundant yet rapid approximation of functions. Video (YouTube)
 16.20-16.35 Zdenek Strakos (Charles University):
The perfidious condition number. Video (YouTube)
 16.40-16.55 Zlatko Drmac (University of Zagreb):
Advances in high accuracy matrix computationsVideo (YouTube)
 17.00-18.00 Reception and poster session
 18.00 Close
Thursday May 30
Session 3. Chair: Sven Hammarling
08.30-09.00 Coffee
09.00-09.25 Ilse Ipsen (North Carolina State University):
Probabilistic roundoff error analysis for fundamental matrix computations
 09.30-09.55 Stefan Güttel (The University of Manchester):
Insights into block rational Krylov methods
10.00-10.25 Françoise Tisseur (The University of Manchester):
Wilkinson’s Improvement Algorithm and Mixed Precision Eigensolvers
10.30-11.00 Coffee and posters
11.00-11.25 Jennifer Pestana (University of Strathclyde):
Preconditioned iterative methods for Toeplitz systems
11.30-11.55 Mawussi Zounon (The Numerical Algorithms Group):
Numerical Linear Algebra Routines for Emerging Computer Architectures
 12.00-12.25 Yuji Nakatsukasa (University of Oxford):
Stable and efficient QR factorization and least-squares solver based on CholeskyQR
12.30-14.00 Lunch and posters
Session 4. Chair: Françoise Tisseur
14.00-14.15 Joao Cardoso (Coimbra Polytechnic-ISEC, Portugal):
Computation of Matrix Gamma Function
 14.20-14.35 Jonathan Cockayne (University of Warwick):
Probabilistic Linear Solvers
14.40-14.55 Lawrence Mulholland (The Numerical Algorithms Group):
Creating an Adjoint Library from the NAG Primal Code Base
15.00-15.15 Vanni Noferini (Aalto University):
Wilkinson’s bus: weak condition numbers, with applications
15.20-15.25 Closing Remarks
15.25 Tea and close

Abstracts are now available.

Group Photo

Here is the group photo from the meeting (hires version).190529_01_small

Registration

The conference is now fully subscribed and registration is closed.

Contributed Talks and Posters

Delegates can submit, during the registration process, abstracts for consideration as a 15 minute talk or poster presentation. A limited number can be accepted because of time and space constraints. For full consideration you should register by April 1, 2019.

Student and Early Career Funding

A limited amount of funding, up to £150 per applicant, is available to support travel and accommodation for PhD students and early career researchers. This can be requested during the registration process. For full consideration you should register by April 1, 2019.

Sponsorship

The Royal Society, The Alan Turing InstituteThe QJMAM Fund for Applied MathematicsThe Numerical Algorithm Group and National Physical Laboratory. 

rslogoTuring Instituteqjmam-with-bordernag_22r42g99b_1npl-logo-blue

Venue

All talks will take place in Frank Adams Room 1 in the Alan Turing Building of the School of Mathematics at the University of Manchester.The building is 20 minutes walk from the city centre and 15 minutes walk from Piccadilly or Oxford Road train stations; see how to get here.

Other Information

The organizers are Sven Hammarling, Nick Higham, and Françoise Tisseur.