为什么RxJava在异步处理时只使用大约10个线程

考虑以下代码，我正在尝试让Observables异步运行。

    try {
        DateTime now = DateTime.now();
        Observable
                .from(map.entrySet()).subscribeOn(Schedulers.from(new ForkJoinPool(Runtime.getRuntime().availableProcessors() * 10)))
                .flatMap(Async.toAsync((Map.Entry<String, Info> entry) -> {
                    // processing work, makes multiple http requests for ref data
                }))
                .doOnCompleted(() -> System.out.println("completed yo...."))
                .doOnError(Throwable::printStackTrace)
                .toList()
                .timeout(1, TimeUnit.MINUTES)
                .toBlocking()
                .single()
                ;
        logger.info(now.toString());
        logger.info(DateTime.now().toString());
        saveToFile(gson.toJson(setForRx));
    } catch (Exception e) {
        e.printStackTrace();
    }

输出显示它使用相同的~10个线程进行处理，我该如何增加？

样本输出：

INFO  2015-06-29 15:11:20,524 [rxjava.ConcurrentRxJava] RxComputationThreadPool-3 
INFO  2015-06-29 15:11:20,526 [rxjava.ConcurrentRxJava] RxComputationThreadPool-6 
INFO  2015-06-29 15:11:20,542 [rxjava.ConcurrentRxJava] RxComputationThreadPool-4 
INFO  2015-06-29 15:11:20,546 [rxjava.ConcurrentRxJava] RxComputationThreadPool-7 
INFO  2015-06-29 15:11:20,571 [rxjava.ConcurrentRxJava] RxComputationThreadPool-2
INFO  2015-06-29 15:11:20,694 [rxjava.ConcurrentRxJava] RxComputationThreadPool-1 
INFO  2015-06-29 15:11:20,920 [rxjava.ConcurrentRxJava] RxComputationThreadPool-8
INFO  2015-06-29 15:11:21,035 [rxjava.ConcurrentRxJava] RxComputationThreadPool-7 
INFO  2015-06-29 15:11:21,039 [rxjava.ConcurrentRxJava] RxComputationThreadPool-4 
INFO  2015-06-29 15:11:21,055 [rxjava.ConcurrentRxJava] RxComputationThreadPool-5
INFO  2015-06-29 15:11:21,081 [rxjava.ConcurrentRxJava] RxComputationThreadPool-3 
INFO  2015-06-29 15:11:21,094 [rxjava.ConcurrentRxJava] RxComputationThreadPool-6 
INFO  2015-06-29 15:11:21,118 [rxjava.ConcurrentRxJava] RxComputationThreadPool-2 

在我的executor版本中，使用Runtime.getRuntime().availableProcessors() * 10，我得到大小为80的池。RxJava可能做到这一点吗？