Programming Parallel Applications in

Cilk (pronounced \silk") is a C-based language for multithreaded parallel programming. Cilk makes it easy to program irregular parallel applications, especially as compared with data-parallel or message-passing programming systems. A Cilk programmer need not worry about protocols and load balancing, which are handled by Cilk's provably eecient runtime system. Many regular and irregular Cilk applications run nearly as fast on one processor as comparable C programs, and they scale well to many processors. 1 Background and goals Cilk is an algorithmic multithreaded language. The philosophy behind Cilk is that a programmer should concentrate on structuring his program to expose parallelism and exploit locality, leaving the runtime system with the responsibility of scheduling the computation to run eeciently on a given platform. Cilk's runtime system takes care of details like load balancing and communication protocols. Unlike other multithreaded languages, however, Cilk is algorithmic in that the runtime system's scheduler guarantees provably eecient and predictable performance. Cilk grew out of theoretical work 1, 5, 6] on the scheduling of multithreaded computations. The basis of Cilk is a provably good scheduling algorithm that has been the cornerstone of Cilk system development. Cilk's provably good scheduler engendered a performance model that accurately predicts the eeciency of a Cilk program using two simple parameters: work and critical-path length 1, 4, 6]. More recent research has included page faults as a measure of locality 2, 3, 12]. The rst implementation of Cilk was a direct descendent of PCM/Threaded-C 10, 12], a C-based package which provided continuation-passing-style threads on Thinking Machines Corporation's Connection Machine Model CM-5 Supercomputer 13] and which used \work-stealing" as a scheduling policy to improve the load balance and locality of the computation. With the addition of a provably good scheduler and the incorporation of many other parallel programming features, the system was rechristened \Cilk-1." Notable among the platforms that supported Cilk-1 was an adaptively parallel and fault-tolerant network-of-workstations implementation, called Cilk-NOW 1, 7]. The next release, Cilk-2, featured full type-checking, supported all of ANSI C in its C-language subset, and ooered call-return semantics for writing multithreaded procedures 14]. Instead of the C