TU Darmstadt / ULB / TUbiblio

Lightweight Requirements Engineering for Exascale Co-design

Calotoiu, Alexandru ; Graf, Alexander ; Hoefler, Torsten ; Lorenz, Daniel ; Wolf, Felix (2018)
Lightweight Requirements Engineering for Exascale Co-design.
2018 IEEE International Conference on Cluster Computing (CLUSTER). Belfast, United Kingdom (10.-13.09.2018)
doi: 10.1109/CLUSTER.2018.00038
Conference or Workshop Item, Bibliographie

Abstract

Given the tremendous cost of an exascale system, its architecture must match the requirements of the applications it is supposed to run as precisely as possible. Conversely, applications must be designed such that building an appropriate system becomes feasible, motivating the idea of co-design. In this process, a fundamental aspect of the application requirements are the rates at which the demands for different resources grow as a code is scaled to a larger machine. However, if the anticipated scale exceeds the size of available platforms this demand can no longer be measured. This is clearly the case when designing an exascale system. Moreover, creating analytical models to predict these requirements is often too laborious - especially when the number and complexity of target applications is high. In this paper, we show how automated performance modeling can be used to quickly predict application requirements for varying scales and problem sizes. Following this approach, we determine the exascale requirements of five scientific codes and use them to illustrate system design tradeoffs.

Item Type: Conference or Workshop Item
Erschienen: 2018
Creators: Calotoiu, Alexandru ; Graf, Alexander ; Hoefler, Torsten ; Lorenz, Daniel ; Wolf, Felix
Type of entry: Bibliographie
Title: Lightweight Requirements Engineering for Exascale Co-design
Language: English
Date: 1 November 2018
Publisher: IEEE
Book Title: Proceedings: 2018 IEEE International Conference on Cluster Computing
Event Title: 2018 IEEE International Conference on Cluster Computing (CLUSTER)
Event Location: Belfast, United Kingdom
Event Dates: 10.-13.09.2018
DOI: 10.1109/CLUSTER.2018.00038
Abstract:

Given the tremendous cost of an exascale system, its architecture must match the requirements of the applications it is supposed to run as precisely as possible. Conversely, applications must be designed such that building an appropriate system becomes feasible, motivating the idea of co-design. In this process, a fundamental aspect of the application requirements are the rates at which the demands for different resources grow as a code is scaled to a larger machine. However, if the anticipated scale exceeds the size of available platforms this demand can no longer be measured. This is clearly the case when designing an exascale system. Moreover, creating analytical models to predict these requirements is often too laborious - especially when the number and complexity of target applications is high. In this paper, we show how automated performance modeling can be used to quickly predict application requirements for varying scales and problem sizes. Following this approach, we determine the exascale requirements of five scientific codes and use them to illustrate system design tradeoffs.

Uncontrolled Keywords: DFG|320898076; BMBF|01IH16008D; DoE|DE-SC0015524
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Parallel Programming
Zentrale Einrichtungen
Zentrale Einrichtungen > University IT-Service and Computing Centre (HRZ)
Zentrale Einrichtungen > University IT-Service and Computing Centre (HRZ) > Hochleistungsrechner
Date Deposited: 10 Jul 2018 14:16
Last Modified: 17 May 2024 08:27
PPN: 51839333X
Export:
Suche nach Titel in: TUfind oder in Google
Send an inquiry Send an inquiry

Options (only for editors)
Show editorial Details Show editorial Details