4-core running Stream's multi-threading is not as good as Stream's single-threading

Above code:

    public static void main(String[] args) {
        System.out.println(String.format("本计算机的核数：%d", Runtime.getRuntime().availableProcessors()));
        // 产⽣100w个随机数(1 ~ 100)，组成列表
        Random random = new Random();
        List<Integer> list = new ArrayList<>(1000_0000);
        for (int i = 0; i < 1000_0000; i++) {
            list.add(random.nextInt(100));
        }
        long prevTime = getCurrentTime();
        list.stream().reduce((a, b) -> a + b).ifPresent(System.out::println);
        System.out.println(String.format("单线程计算耗时：%d", getCurrentTime() - prevTime ));
        prevTime = getCurrentTime();
        list.stream().parallel().reduce((a, b) -> a + b).ifPresent(System.out::println);
        System.out.println(String.format("多线程计算耗时：%d", getCurrentTime() -prevTime));
    }

result:

The number of cores of this computer: 4
495028607
Time-consuming single-thread calculation: 227
495028607
Time-consuming multi-thread calculation: 979

The multi-threading of Stream running on 4 cores is not as good as the single-threading of Stream!