93 lines
2.2 KiB
Go
93 lines
2.2 KiB
Go
|
package main
|
||
|
|
|
||
|
|
import (
|
||
|
|
"fmt"
|
||
|
|
"math/rand"
|
||
|
|
"sync"
|
||
|
|
"time"
|
||
|
|
)
|
||
|
|
|
||
|
|
func sequentialMatrixMultiply(matrixA, matrixB [][]float64) [][]float64 {
|
||
|
|
rowsA, colsA := len(matrixA), len(matrixA[0])
|
||
|
|
colsB := len(matrixB[0])
|
||
|
|
result := make([][]float64, rowsA)
|
||
|
|
for i := range result {
|
||
|
|
result[i] = make([]float64, colsB)
|
||
|
|
}
|
||
|
|
|
||
|
|
for i := 0; i < rowsA; i++ {
|
||
|
|
for j := 0; j < colsB; j++ {
|
||
|
|
for k := 0; k < colsA; k++ {
|
||
|
|
result[i][j] += matrixA[i][k] * matrixB[k][j]
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}
|
||
|
|
return result
|
||
|
|
}
|
||
|
|
|
||
|
|
func parallelMatrixMultiply(matrixA, matrixB [][]float64, numProcesses int) [][]float64 {
|
||
|
|
rowsA, colsA := len(matrixA), len(matrixA[0])
|
||
|
|
colsB := len(matrixB[0])
|
||
|
|
result := make([][]float64, rowsA)
|
||
|
|
for i := range result {
|
||
|
|
result[i] = make([]float64, colsB)
|
||
|
|
}
|
||
|
|
|
||
|
|
var wg sync.WaitGroup
|
||
|
|
wg.Add(numProcesses)
|
||
|
|
|
||
|
|
for i := 0; i < numProcesses; i++ {
|
||
|
|
go func(id, startRow, endRow int) {
|
||
|
|
defer wg.Done()
|
||
|
|
for i := startRow; i < endRow; i++ {
|
||
|
|
for j := 0; j < colsB; j++ {
|
||
|
|
for k := 0; k < colsA; k++ {
|
||
|
|
result[i][j] += matrixA[i][k] * matrixB[k][j]
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}
|
||
|
|
}(i, i*rowsA/numProcesses, (i+1)*rowsA/numProcesses)
|
||
|
|
}
|
||
|
|
|
||
|
|
wg.Wait()
|
||
|
|
return result
|
||
|
|
}
|
||
|
|
|
||
|
|
func runTest(matrixSize, numProcesses int) {
|
||
|
|
matrixA := generateRandomMatrix(matrixSize, matrixSize)
|
||
|
|
matrixB := generateRandomMatrix(matrixSize, matrixSize)
|
||
|
|
|
||
|
|
startTime := time.Now()
|
||
|
|
_ = sequentialMatrixMultiply(matrixA, matrixB)
|
||
|
|
sequentialTime := time.Since(startTime)
|
||
|
|
fmt.Printf("Sequential matrix multiplication took (%dx%d): %s\n", matrixSize, matrixSize, sequentialTime)
|
||
|
|
|
||
|
|
startTime = time.Now()
|
||
|
|
_ = parallelMatrixMultiply(matrixA, matrixB, numProcesses)
|
||
|
|
parallelTime := time.Since(startTime)
|
||
|
|
fmt.Printf("Parallel matrix multiplication with %d threads took (%dx%d): %s\n", numProcesses, matrixSize, matrixSize, parallelTime)
|
||
|
|
}
|
||
|
|
|
||
|
|
func generateRandomMatrix(rows, cols int) [][]float64 {
|
||
|
|
matrix := make([][]float64, rows)
|
||
|
|
for i := range matrix {
|
||
|
|
matrix[i] = make([]float64, cols)
|
||
|
|
for j := range matrix[i] {
|
||
|
|
matrix[i][j] = rand.Float64()
|
||
|
|
}
|
||
|
|
}
|
||
|
|
return matrix
|
||
|
|
}
|
||
|
|
|
||
|
|
func main() {
|
||
|
|
rand.Seed(time.Now().UnixNano())
|
||
|
|
|
||
|
|
// Benchmarks for matrices with sizes 100x100, 300x300, and 500x500 with different numbers of processes
|
||
|
|
runTest(100, 2)
|
||
|
|
runTest(100, 4)
|
||
|
|
runTest(300, 2)
|
||
|
|
runTest(300, 4)
|
||
|
|
runTest(500, 2)
|
||
|
|
runTest(500, 4)
|
||
|
|
}
|