In many ways, the history of programming languages has been driven by, and certainly intertwined with, the needs of numerical and scientific computing. The first high-level programming language, Fortran, was created to solve scientific computing problems, and continues to be important in the field even to this day. In recent years, the rise of data science as a specialty has brought additional focus to numerical computing, particularly for statistical uses. In this area, somewhat counter-intuitively, both specialized languages such as R and general-purpose languages such as Python are in widespread use. The rise of Hadoop and Spark has spread the use of Java and Scala respectively among this community. In the midst of all this, Matlab has had a strong niche within engineering communities, while Mathematica remains unparalleled for symbolic operations.
A new language for scientific computing therefore has a very high barrier to overcome, and it's been only a few short years since the Julia language was introduced to the world. In that time, however, its innovative features, combining the ease of use of a dynamic language and the performance of a statically compiled language, have created a growing niche within the numerical computing world. Based on multiple dispatch as its defining paradigm, Julia is a very pleasant language to program in, making mathematical abstractions very easy to express. However, it was the claim of high performance that drew the earliest adopters.
This, then, is a book that celebrates writing high-performance programs. With Julia, this is not only possible, but also reasonably straightforward, in a low-overhead, dynamic language.
As a reader of this book, you have likely already written your first few Julia programs. We will assume that you have successfully installed Julia, and have a working programming environment available. We expect you are familiar with very basic Julia syntax, but we will discuss and review many of those concepts throughout the book as we introduce them.
In this chapter, we will describe some of the underlying design elements of Julia that contribute to its well-deserved reputation as a fast language:
- Julia – fast and dynamic
- Designed for speed
- How fast can Julia be?
