WEBVTT 00:01.140 --> 00:01.710 Hi everyone. 00:01.710 --> 00:03.030 Welcome back to this tutorial. 00:03.030 --> 00:09.930 In this tutorial we will code a simple example on when not to use parallel streams. 00:09.930 --> 00:15.420 I'm going to create a class in the parallel streams package the class name is going to be. 00:18.340 --> 00:24.820 Parallel stream boxed example. 00:25.420 --> 00:30.100 Then I'm going to make this class executable by adding public static void main method. 00:30.910 --> 00:31.720 And then. 00:34.010 --> 00:38.810 I'm going to create a list of integers instead of using a int stream. 00:38.810 --> 00:45.500 I'm going to use a list of integers, but I'm going to use the int stream to create that. 00:47.060 --> 00:52.310 INT stream dot range closed one comma. 00:54.200 --> 00:54.950 10,000. 00:55.310 --> 01:01.400 And then what I'm going to do, I'm going to do a boxed and then collectors. 01:02.990 --> 01:15.950 It's going to be to list you do this, it is going to give you a list of integers, basically wrapper 01:15.950 --> 01:20.780 classes, a list of 10,000 integer elements. 01:20.780 --> 01:24.020 But these are not primitive integers. 01:24.020 --> 01:28.160 These are wrapper integer objects. 01:28.160 --> 01:31.340 And then now I'm going to create a. 01:33.020 --> 01:38.750 Method called sequential sum. 01:41.210 --> 01:52.750 This is going to accept this list of integers as an input, and then this is going to perform the summation 01:52.760 --> 01:55.760 integer list, dot stream. 01:57.090 --> 01:57.530 Dot. 01:57.540 --> 02:04.020 I'm going to use a reduce function, which is going to take the initial value as zero, and then it 02:04.020 --> 02:11.460 is going to accept the binary operator, which is going to take two input as an input and then perform 02:11.460 --> 02:15.630 the summation of this two input and return the result. 02:17.190 --> 02:18.720 Basically this is. 02:19.850 --> 02:24.590 Nothing but performing summation of all the elements. 02:26.670 --> 02:32.280 Of the list of integer type with the name integer list. 02:33.270 --> 02:35.610 Now I'm going to perform the 02:38.310 --> 02:42.600 calculation of how long did it take in order to complete this process? 02:42.780 --> 02:45.540 System dot current milliseconds. 02:45.840 --> 02:48.510 And then I'm going to have the. 02:49.960 --> 02:55.690 Duration equal to system.currenttimemillis seconds minus the start. 02:57.630 --> 03:07.090 So I'm going to have this value as duration in sequential stream. 03:09.530 --> 03:12.980 Duration and then the return. 03:14.150 --> 03:15.560 Is equal to some. 03:18.650 --> 03:20.310 I'm going to create another method. 03:20.330 --> 03:24.200 It's going to be named as parallel sum. 03:27.390 --> 03:32.400 And in the place of cream, I'm going to use parallel stream. 03:32.400 --> 03:34.470 So now we have two methods. 03:34.470 --> 03:40.320 One is performing the sequential stream and another one is performing the parallel stream on the input 03:40.350 --> 03:42.660 of type list of integers. 03:42.690 --> 03:51.870 Let's go ahead and run this and then check what is the performance that it is going to value the duration. 03:53.070 --> 03:59.820 How long did it take in order to perform the summation of 10,000 elements in the parallel stream and 03:59.820 --> 04:01.380 the sequential stream? 04:05.020 --> 04:06.010 To do list 04:08.590 --> 04:11.990 and this is going to be parallel sum. 04:12.010 --> 04:12.760 There we go. 04:12.760 --> 04:13.930 So we have the code ready. 04:13.930 --> 04:15.780 Let's run this and check the result. 04:15.790 --> 04:19.300 How long did it take for the sequential stream to perform the summation? 04:19.300 --> 04:25.060 How long did it take for the parallel stream to perform the summation of 10,000 elements? 04:25.390 --> 04:31.750 If we take a look at it, the sequential stream took only four milliseconds here and the parallel stream 04:31.780 --> 04:34.120 took eight milliseconds. 04:34.120 --> 04:40.090 The reason here is if we take a look at the input type is of list of integers. 04:40.090 --> 04:48.790 So here there is an effort which is needed in order to perform the unboxing right. 04:48.790 --> 04:57.430 So it needs to convert unboxing from integer to int. 04:57.460 --> 05:02.530 So it needs to perform unboxing behind the scenes for each and every value. 05:02.530 --> 05:07.760 This parallel stream passes to it and then returns the result. 05:08.150 --> 05:14.870 It is always recommended to check the performance before going to the assumption that parallel stream 05:14.870 --> 05:16.280 is going to perform better. 05:16.280 --> 05:18.530 So we have looked at an example. 05:18.530 --> 05:24.680 So there is going to be many more examples which might look like parallel stream is going to perform 05:24.680 --> 05:32.180 better, but in reality parallel stream is not going to perform better than the sequential stream in 05:32.180 --> 05:33.620 all the scenarios. 05:34.360 --> 05:34.750 But this. 05:34.750 --> 05:36.430 We came to the end of this tutorial. 05:36.460 --> 05:37.720 Thank you for watching.