Matrix multiplication with parallel collections

gistfile1.scala

  override def multiply(m1: Array[Array[Double]], m2: Array[Array[Double]]) :  Array[Array[Double]] = {
    val res =  Array.ofDim[Double](m1.length, m2(0).length)
    val M1_COLS = m1(0).length
    val M1_ROWS = m1.length
    val M2_COLS = m2(0).length

    @inline def singleThreadedMultiplicationFAST(start_row:Int,  end_row:Int) {
      var col, i  = 0
      var sum = 0.0
      var row = start_row

      // while statements are much faster than for statements
      while(row < end_row){ col = 0
        while(col < M2_COLS){ i = 0; sum = 0
          while(i<M1_COLS){
            sum += m1(row)(i) * m2(i)(col)
            i+=1
          }

          res(row)(col) = sum
          col += 1

        }; row += 1
      }
    }

    (0 until M1_ROWS).par.foreach( i =>
      singleThreadedMultiplicationFAST(i, i+1)
    )

    res

  }