FPBench Logo

FPTalks 2022

The leading edge of floating-point research

The floating-point research community.
Online. July 6, 2022.

FPTalks 2022 was held online on July 6, 2022 on Zoom. All talks are published on Youtube, with individual talks accessible via the links below. Talks were be 10 minutes long, followed by audience questions over Slack.

Session 1

Welcome

Zachary Tatlock, University of Washington & Bill Zorn, Intel

The hardware costs of Posits and IEEE floating-point

Florent de Dinechin, INSA Lyon

Correctly rounded power function in double precision

Tom Hubrecht, CERN, INRIA, ENS

Compressed Matrix Computations

Matthieu Martel, University of Perpignan

Break

Session 2

Correctly Rounded Elementary Functions with the RLIBM Approach

Santosh Nagarakatte, Rutgers University

FPVM: Towards a Floating Point Virtual Machine

Peter Dinda, Northwestern University

VSCode-PRECISA: a Toolkit for Floating-Point Round-Off Error Analysis

Paolo Masci, NIA/NASA LaRC

Design Space Exploration for Polynomial Interpolation Hardware

Theo Drane, Intel

Break

Session 3

Accelerating scientific computing applications with TPUs

Tianjian Lu, Google Research

Visualization of Parameter and Floating-Point Precision Level Changes for ODE Solvers

Mike Lam, James Madison University

An Exceptional FP Topic

Xinyi Li, University of Utah

What every 21st century computer scientist should know about floating-point

Bill Zorn, Intel