Industrial Applications of High-Performance Computing

1

Chapter 1

Introduction and

Brief History of

Supercomputing

Merle Giles and Anwar Osseyran

1.1 SUPERCOMPUTING 101

It should be noted that supercomputing, high-performance computing

(HPC),and advanced computing are terms that are oen used interchange-

ably. ese advanced machines use components similar to everyday

computers, such as memory, processors known as CPUs (central processing

units), datastorage, communication, and soware, but supercomputers use

these in ways that produce speedy results. e “trick” to supercomputing is

in the scale of use by stringing together vast numbers of each of the compo-

nents listed above and parallelizing data movement at every possible step.

Typically, communication between multiple computers in a data center

occurs using common Ethernet cables, oen seen in the form of orange

CONTENTS

1.1 Supercomputing 101 1

1.2 e Dawn of Digital Electronics: 1930s–1940s 4

1.3 e Dawn of Supercomputing: 1950s–1970s 8

1.4 Growth and Commoditization of Supercomputing: 1980s–2000s 10

1.5 Supercomputers for Open Science and Open Engineering 12

1.6 Supercomputers for Industry 16

1.7 Modern Industrial Supercomputing 18

Additional Reading 20

References 21

2 ◾ Industrial Applications of High-Performance Computing

or blue cables in homes and businesses. In supercomputing applications,

even faster communication is used and highly proprietary communication

schemes push speeds to 4–100 times the speeds of common Ethernet,

using parallel, simultaneous data paths to push and pull data from CPUs

to storage and memory and back.

Four basic components of supercomputing can be described as follows:

1. Parallel processing, using multiple cores per CPU, multiple CPUs in a

single computer, and multiple computers in a supercomputer. Agood

consumer CPU example is a dual-core or quad-core processor in a

laptop computer, which allows soware applications to do two or

four operations at once. CPUs used in supercomputers will typically

have between 8 and 32 cores, with newer general purpose graphics

processing units (GPUs) deploying hundreds of cores.

2. Large memory, with each computer holding 32, 64, 256, or

more gigabytes of memory, depending on the specic science or

engineering need.

3. Multiprocessor, or multinode communication, allowing an entire

collection of computers to be used in a single application. Special

soware techniques, such as MPI (Message Passing Interface) must

be used for sharing an application across two or more computers.

Supercomputers might pass information simultaneously among

25,000 or more individual computers.

4. Parallel le systems move data between compute nodes and storage

(disks, magnetic tape, and solid-state devices), using multiple

channels simultaneously, rather than one, to speed up data input

and output. In very advanced systems, large caches of memory sit

between the storage and compute systems to minimize bottlenecks.

As in all computers, supercomputers require an operating system—

commonly Linux—as well as machine-management soware known

as middleware. Specic application soware will map to the operating

system, and ultimately to the machine components themselves through

the use of compilers, which translate word-based programs into machine

language that only recognizes zeros and ones.

Supercomputing adds signicant complexity for users due to

idiosyncrasies of various vendor platforms, processors, storage disks, and

so on. ese dierences are routinely subtle and nonobvious, leading to