Parallel Processing with Haskell

Download Haskell now. It's a pure functional language with built-in support for parallel processing. That means it is more Lispy than Lisp and faster than C.

• maths.hs

import Text.Printf
import Control.Parallel

main = (a::Integer) `par` (b::Integer) `par` (c::Integer) `pseq`
	(printf "A = %d\nB = %d\nC = %d\n" a b c) where
	a = ack 3 10
	b = fac 42
	c = fib 34

fac 0 = 1
fac n = n * fac (n-1)

ack 0 n = n + 1
ack m 0 = ack (m-1) 1
ack m n = ack (m-1) (ack m (n-1))

fib 0 = 0
fib 1 = 1
fib n = fib (n-1) + fib (n-2)

Compile

$ ghc -O2 --make maths.hs -threaded -rtsopts
[1 of 1] Compiling Main             ( maths.hs, maths.o )
Linking maths ...

Run

$ time ./maths +RTS -N
A = 8189
B = 1405006117752879898543142606244511569936384000000000
C = 5702887

real	0m0.774s
user	0m1.201s
sys	0m0.014s

Debriefing

This program computes the ackermann, factorial, and fibonacci functions in parallel. par instructs Haskell to calculate a, b, and c independently. pseq instructs Haskell to calculate the values before printing them.

Compilation options

• --make compiles the binary
• -O2 optimizes the binary
• -threaded provides the parallel processing feature

Runtime options

• time runs our Haskell program and prints the time it requires to run.
• +RTS installs signals capabilities, so that we can cancel the program with Control+C if we have to.
• -Ncores specifies multicore, with an optional manual specification of the number of cores. If you specify more cores than you actually have, the program stalls, and the +RTS option becomes crucial.

So there you have it. A parallel processing program with no threading headaches. Haskell does the grunt work for you so you can be more a productive programmer.

Resources

Haskell.org is the official Haskell website.

Try Haskell is an online, interactive tutorial for learning Haskell.

Learn You a Haskell is a funny tutorial explaining list comprehensions.

Haskell in 5 steps inspired this tutorial.