eduardlopez/quicksort-parallel

Quicksort algorithm implemented in OpenMP

quicksort-parallel

Parallel quicksort algorithm implemented in OpenMP

##Execution demo: Some technical info of the computer:

Processor	Intel Core i7 950 (4 Cores x 2 Threads per core = 8 Hardware Threads)
Processor frequency	Base:3.06Ghz, Turbo:3.33GHz
DRAM	6 GBytes
Cache	L3:8MB, L2:256KB, L1:32KB
OS	Scientific Linux release 6.7 (Carbon)
Compiler	gcc version 4.9.0 (GCC)

####Compilation: ``` gcc -O3 -fopenmp quicksort-demo-comparison.c -o quicksort-demo-comparison ``` ####Execution: ``` ./quicksort-demo-comparison ``` ####Output Result: ``` Size of the array (aprox.): 1000 MBytes TOTAL MEMORY ALLOCATED: 3 x array = (aprox.): 3000 MBytes

Initializing the arrays with random numbers... Initialization complete

Sorting with serial 'qsort' function of 'stdlib.h' ... Sorted in (aprox.): 45.110000 seconds

Sorting with custom serial QuickSort... Sorted in (aprox.): 33.530000 seconds

Sorting with custom PARALLEL QuickSort... Sorted in (aprox.): 6.699294 seconds

Checking if the results are correct... The result with 'custom serial QuickSort' is CORRECT The result with 'custom PARALLEL QuickSort' is CORRECT

quicksort-parallel's Issues

Slower than qsort on small arrays

I have tested the code on smaller arrays ~5000 elements and it is significantly slower than the sequential version.