The Challenge
The exponentially increasing need for computing power in science and society at large has fueled the quest for developing exascale computers capable of performing a billion billion (10^18) floating-point operations per second. This upcoming generation of high performance computing (HPC) will rely on an intricate interplay between thousands of sophisticated processing nodes, each with a large number of cores, deep memory hierarchies and equipped with accelerators, organized in complex communication topologies. Heterogeneity is expected on all hardware levels, such as the nodes may also include processors that are tailored for certain types of algorithms, for instance graph-oriented methods for machine learning.
While there is no fixed blueprint for exascale computers, the aggregated level of complexity in a strongly heterogeneous system designed for billion-way concurrency also represents a major challenge with many features: How to program such computers? How to port existing code, and how to reach a satisfactory level of reliability and efficiency while maintaining an acceptable energy footprint? How do you even debug software running on such a beast? While competing technologies in hardware, middleware, and software are being driven by major research projects in the United States, China, Japan and the EU, it is essential for Norwegian HPC research groups to keep pace with the frontier research.
The eX3 project
Together with partners, Simula Research Laboratory has established eX3 to become the previously missing national resource that can prepare researchers for exascale computing in Norway. The eX3 infrastructure is currently funded for five years through the RCN program for national research infrastructures. In addition to the host institution Simula, the project consortium also counts the national HPC management body Sigma2, HPC research groups from the University of Tromsø, NTNU, the University of Bergen, and OsloMet, as well as the HPC technology providers Graphcore, Dolphin Interconnect Solutions, and Numascale.
The eX3 infrastructure is not an exascale computer by itself, but it is a carefully curated ecosystem of technology components that will be crucial for embracing exascale computing. It will allow HPC researchers throughout Norway and their collaborators to experiment hands-on with emerging HPC technologies – hardware as well as software.