Concurrent computing is a form of computing in which several computations are executed concurrently—during overlapping time periods—instead of sequentially—with one completing before the next starts.
This is a property of a system—whether a program, computer, or a network—where there is a separate execution point or "thread of control" for each process. A concurrent system is one where a computation can advance without waiting for all other computations to complete.Operating System Concepts 9th edition, Abraham Silberschatz. "Chapter 4: Threads"
Concurrent computing is a form of modular programming. In its paradigm an overall computation is factored into subcomputations that may be executed concurrently. Pioneers in the field of concurrent computing include Edsger Dijkstra, Per Brinch Hansen, and C.A.R. Hoare.
Related concepts Concurrency is a broader concept that encompasses several related ideas, including:
Parallelism (simultaneous execution on multiple processing units). Parallelism executes tasks independently on multiple CPU cores. Concurrency allows for multiple threads of control at the program level, which can use parallelism or time-slicing to perform these tasks. Programs may exhibit parallelism only, concurrency only, both parallelism and concurrency, or neither. Multi-threading and multi-processing (shared system resources) Synchronization (coordinating access to shared resources) Coordination (managing interactions between concurrent tasks) Concurrency Control (ensuring data consistency and integrity) Inter-process Communication (IPC, facilitating information exchange)
Introduction
Concurrency refers to the ability of a system to execute multiple tasks through simultaneous execution or time-sharing (context switching), sharing resources and managing interactions. Concurrency improves responsiveness, throughput, and scalability in modern computing, including: Operating systems and embedded systems Distributed systems, parallel computing, and high-performance computing Database systems, web applications, and cloud computing
The concept of concurrent computing is frequently confused with the related but distinct concept of parallel computing,Pike, Rob (2012-01-11). "Concurrency is not Parallelism". Waza conference, 11 January 2012. Retrieved from http://talks.golang.org/2012/waza.slide (slides) and http://vimeo.com/49718712 (video). although both can be described as "multiple processes executing during the same period of time". In parallel computing, execution occurs at the same physical instant: for example, on separate processors of a multi-processor machine, with the goal of speeding up computations—parallel computing is impossible on a (one-core) single processor, as only one computation can occur at any instant (during any single clock cycle). By contrast, concurrent computing consists of process lifetimes overlapping, but execution does not happen at the same instant. The goal here is to model processes that happen concurrently, like multiple clients accessing a server at the same time. Structuring software systems as composed of multiple concurrent, communicating parts can be useful for tackling complexity, regardless of whether the parts can be executed in parallel.
For example, concurrent processes can be executed on one core by interleaving the execution steps of each process via time-sharing slices: only one process runs at a time, and if it does not complete during its time slice, it is paused, another process begins or resumes, and then later the original process is resumed. In this way, multiple processes are part-way through execution at a single instant, but only one process is being executed at that instant.
Concurrent computations may be executed in parallel, for example, by assigning each process to a separate processor or processor core, or distributing a computation across a network.
The exact timing of when tasks in a concurrent system are executed depends on the scheduling, and tasks need not always be executed concurrently. For example, given two tasks, T1 and T2:
T1 may be executed and finished before T2 or vice versa (serial and sequential) T1 and T2 may be executed alternately (serial and concurrent) T1 and T2 may be executed simultaneously at the same instant of time (parallel and concurrent)
The word "sequential" is used as an antonym for both "concurrent" and "parallel"; when these are explicitly distinguished, concurrent/sequential and parallel/serial are used as opposing pairs. A schedule in which tasks execute one at a time (serially, no parallelism), without interleaving (sequentially, no concurrency: no task begins until the prior task ends) is called a serial schedule. A set of tasks that can be scheduled serially is serializable, which simplifies concurrency control.
Coordinating access to shared resources The main challenge in designing concurrent programs is concurrency control: ensuring the correct sequencing of the interactions or communications between different computational executions, and coordinating access to resources that are shared among executions. Potential problems include race conditions, deadlocks, and resource starvation. For example, consider the following algorithm to make withdrawals from a checking account represented by the shared resource balance:
bool withdraw(int withdrawal) { if (balance >= withdrawal)
balance -= withdrawal; return true;
return false; }
Suppose balance = 500, and two concurrent threads make the calls withdraw(300) and withdraw(350). If line 3 in both operations executes before line 5 both operations will find that balance >= withdrawal evaluates to true, and execution will proceed to subtracting the withdrawal amount. However, since both processes perform their withdrawals, the total amount withdrawn will end up being more than the original balance. These sorts of problems with shared resources benefit from the use of concurrency control, or non-blocking algorithms.
Advantages There are advantages of concurrent computing:
Increased program throughput—parallel execution of a concurrent algorithm allows the number of tasks completed in a given time to increase proportionally to the number of processors according to Gustafson's law. High responsiveness for input/output—input/output-intensive programs mostly wait for input or output operations to complete. Concurrent programming allows the time that would be spent waiting to be used for another task. More appropriate program structure—some problems and problem domains are well-suited to representation as concurrent tasks or processes. For example MVCC.
Issues Because computations in a concurrent system can interact with each other while being executed, the number of possible execution paths in the system can be extremely large, and the resulting outcome can be indeterminate. Concurrent use of shared resources can be a source of indeterminacy, leading to issues such as deadlocks, and resource starvation.
Design of concurrent systems often entails finding reliable techniques for coordinating their execution, data exchange, memory allocation, and execution scheduling to minimize response time and maximize throughput.
Theory Concurrency theory has been an active field of research in theoretical computer science. One of the first proposals was Carl Adam Petri's seminal work on Petri nets in the early 1960s. In the years since, a wide variety of formalisms have been developed for modeling and reasoning about concurrency.
Models A number of formalisms for modeling and understanding concurrent systems have been developed, including: The parallel random-access machine The actor model Computational bridging models such as the bulk synchronous parallel (BSP) model Petri nets Process calculi Calculus of communicating systems (CCS) Communicating sequential processes (CSP) model π-calculus Tuple spaces, e.g., Linda Simple Concurrent Object-Oriented Programming (SCOOP) Reo Coordination Language Trace monoids
Some of these models of concurrency are primarily intended to support reasoning and specification, while others can be used through the entire development cycle, including design, implementation, proof, testing, and simulation of concurrent systems. Some of these are based on message passing, while others have different mechanisms for concurrency.
The proliferation of different models of concurrency has motivated some researchers to develop ways to unify these different theoretical models. For example, Lee and Sangiovanni-Vincentelli have demonstrated that a so-called "tagged-signal" model can be used to provide a common framework for defining the denotational semantics of a variety of different models of concurrency, while Nielsen, Sassone, and Winskel have demonstrated that category theory can be used to provide a similar unified understanding of different models.
The Concurrency Representation Theorem in the actor model provides a fairly general way to represent concurrent systems that are closed in the sense that they do not receive communications from outside. (Other concurrency systems, e.g., process calculi can be modeled in the actor model using a two-phase commit protocol.Frederick Knabe. A Distributed Protocol for Channel-Based Communication with Choice PARLE 1992.) The mathematical denotation denoted by a closed system is constructed increasingly better approximations from an initial behavior called using a behavior approximating function to construct a denotation (meaning) for as follows:
In this way, can be mathematically characterized in terms of all its possible behaviors.
Logics Various types of temporal logic can be used to help reason about concurrent systems. Some of these logics, such as linear temporal logic and computation tree logic, allow assertions to be made about the sequences of states that a concurrent system can pass through. Others, such as action computational tree logic, Hennessy–Milner logic, and Lamport's temporal logic of actions, build their assertions from sequences of actions (changes in state). The principal application of these logics is in writing specifications for concurrent systems.
Models Introduced in 1962, Petri nets were an early attempt to codify the rules of concurrent execution. Dataflow theory later built upon these, and Dataflow architectures were created to physically implement the ideas of dataflow theory. Beginning in the late 1970s, process calculi such as Calculus of Communicating Systems (CCS) and Communicating Sequential Processes (CSP) were developed to permit algebraic reasoning about systems composed of interacting components. The π-calculus added the capability for reasoning about dynamic topologies.
Input/output automata were introduced in 1987.
Logics such as Lamport's TLA+, and mathematical models such as traces and Actor event diagrams, have also been developed to describe the behavior of concurrent systems.
Software transactional memory borrows from database theory the concept of atomic transactions and applies them to memory accesses.
Consistency models
Concurrent programming languages and multiprocessor programs must have a consistency model (also known as a memory model). The consistency model defines rules for how operations on computer memory occur and how results are produced.
One of the first consistency models was Leslie Lamport's sequential consistency model. Sequential consistency is the property of a program that its execution produces the same results as a sequential program. Specifically, a program is sequentially consistent if "the results of any execution is the same as if the operations of all the processors were executed in some sequential order, and the operations of each individual processor appear in this sequence in the order specified by its program".
Implementation