YoVDO

Numerical Weather Prediction - Facing the Future with C++

Offered By: CppNow via YouTube

Tags

C++Now Courses C++ Courses Meteorology Courses Functional Programming Courses High Performance Computing Courses Parallel Processing Courses

Course Description

Overview

Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore a comprehensive keynote address from C++Now 2015 on Numerical Weather Prediction (NWP) and its future with C++. Delve into the challenges faced by the NWP community, including data-intensive processing, tight time constraints, and the need to modernize software techniques for many-core era hardware. Learn about the transition from monolithic Fortran code to embracing modern software practices, with a focus on high-performance languages like C++ and Domain-Specific Languages (DSLs). Discover how the European Centre for Medium-Range Weather Forecasts (ECMWF) is addressing these challenges, including the development of a C++ Domain-Specific Embedded Language (DSEL) inspired by functional programming for building parallel processing pipelines for meteorological data. Gain insights into the future of NWP software development, including power limitations, data I/O rates, and exponential volume growth, as well as the benefits and challenges of implementing C++ in scientific computing environments.

Syllabus

Intro
Overview
Who are we and what do we do?
A basic description of our models
Major assimilated datasets
4D-Var Assimilation
Data Assimilation
Model Levels
Operational model grid
Grid Effect
Meteorological Fields
Chaos and weather prediction
RMDCN Connections
New Algorithms Challenge
Evalution of ECMWF scores comparison northern and southern hemispheres
Benefits of High Resolution
Planned Resolution Upgrades
Power Challenge
Platform Uncertainty Challenge
Power Matters
Hypothetical Solution
HPC Platform Restrictions
C++ Support on HPC
Scientists are reluctant to use C++
Big Data Challenge
Vis for Volume: Observations
Vis for Volume: Archive
CPU Power Growth
CPU Performance Growth (single-threaded)
Storage Density Growth Multiple Technologies
HDD Storage Growth
What does it imply?
Meteorological Archival and Retrieval System
MARS 2011 Migration
A meteorological language
Current IFS Model
Atlas capabilities
So what about Fortran?
C++ contains implementation
Fortran/C interface
Fortran Program
A look at the data-chain
Observations & Fields: Some similarities
Requirements
Use Case: Product Generation
Use Case: Observation Filters
Use Case Workstation Interpolation & Visualisation
Hermes


Taught by

CppNow

Related Courses

High Performance Computing
Georgia Institute of Technology via Udacity
Введение в параллельное программирование с использованием OpenMP и MPI
Tomsk State University via Coursera
High Performance Computing in the Cloud
Dublin City University via FutureLearn
Production Machine Learning Systems
Google Cloud via Coursera
LAFF-On Programming for High Performance
The University of Texas at Austin via edX