Cray XC40

Cray XC40 is a high-performance computing system developed by Cray Inc., a leading manufacturer of supercomputers used by organizations such as NASA, Los Alamos National Laboratory, and European Organization for Nuclear Research. The Cray XC40 is designed to support a wide range of applications, including climate modeling, materials science, and genomics research, which are critical to the work of institutions like Massachusetts Institute of Technology, Stanford University, and University of California, Berkeley. This system is also used by researchers at Harvard University, University of Oxford, and University of Cambridge to advance our understanding of complex phenomena, such as those studied in particle physics and astrophysics.

Introduction

The Cray XC40 is a scalable, high-performance computing system that is part of the Cray XC series, which also includes the Cray XC30 and Cray XC50. It is designed to support a wide range of applications, including scientific simulations, data analytics, and artificial intelligence, which are used by organizations such as Google, Amazon, and Microsoft. The system is built using Intel Xeon processors, NVIDIA Tesla accelerators, and Cray's proprietary Aries interconnect, which provides high-bandwidth, low-latency communication between nodes, similar to the InfiniBand technology used in high-performance computing systems. The Cray XC40 is also compatible with a range of operating systems, including SUSE Linux Enterprise Server, Red Hat Enterprise Linux, and Ubuntu, which are widely used in the high-performance computing community.

Architecture

The Cray XC40 system is based on a modular architecture, with each node consisting of a blade server that contains multiple Intel Xeon processors, memory, and storage. The nodes are connected using Cray's Aries interconnect, which provides a high-bandwidth, low-latency network for data transfer between nodes, similar to the Ethernet networks used in data centers by companies like Facebook, Apple, and IBM. The system also includes a range of storage options, including hard disk drives, solid-state drives, and flash storage, which are used by organizations such as Netflix, Twitter, and Dropbox to store and manage large amounts of data. The Cray XC40 is designed to be highly scalable, with systems ranging from a few nodes to thousands of nodes, making it suitable for a wide range of applications, from small-scale simulations to large-scale data analytics, which are used by researchers at University of Chicago, University of Michigan, and University of California, Los Angeles.

Performance

The Cray XC40 system is designed to deliver high performance and scalability, with a range of features that optimize its performance for different types of applications, such as climate modeling and materials science, which are critical to the work of institutions like National Center for Atmospheric Research and Lawrence Livermore National Laboratory. The system's Aries interconnect provides high-bandwidth, low-latency communication between nodes, which is essential for applications that require high levels of parallelism, such as scientific simulations and data analytics, which are used by companies like Palantir Technologies and SAP SE. The system also includes a range of optimization tools and libraries, such as Cray's LibSci library, which provides optimized implementations of common scientific algorithms, similar to the NumPy library used in Python programming. The Cray XC40 has been used to achieve high performance on a range of benchmarks, including LINPACK and HPL-AI, which are widely used in the high-performance computing community to evaluate the performance of systems like the Cray XC40.

Applications

The Cray XC40 system is designed to support a wide range of applications, including scientific simulations, data analytics, and artificial intelligence, which are used by organizations such as National Institutes of Health, National Science Foundation, and European Space Agency. The system is particularly well-suited to applications that require high levels of parallelism, such as climate modeling, materials science, and genomics research, which are critical to the work of researchers at University of California, San Diego, University of Illinois at Urbana-Champaign, and University of Wisconsin-Madison. The Cray XC40 is also used in a range of industries, including finance, healthcare, and energy, where high-performance computing is used to analyze large amounts of data and simulate complex systems, similar to the simulations used in video games and movie production by companies like Electronic Arts and Industrial Light & Magic.

History

The Cray XC40 system was announced in 2014, as part of Cray's Cray XC series, which also includes the Cray XC30 and Cray XC50. The system was designed to provide a high-performance, scalable platform for a wide range of applications, from small-scale simulations to large-scale data analytics, which are used by researchers at University of Texas at Austin, University of Washington, and University of Pennsylvania. The Cray XC40 has been used by a range of organizations, including NASA, Los Alamos National Laboratory, and European Organization for Nuclear Research, to advance our understanding of complex phenomena, such as those studied in particle physics and astrophysics. The system has also been used to achieve a range of milestones, including petascale performance and exaflop scaling, which are critical to the development of future supercomputers by companies like Intel Corporation and NVIDIA Corporation. Category:Supercomputers