Github user cloud-fan commented on the issue: https://github.com/apache/spark/pull/19222 Instead of round-robin the memory block type across iterations, can we just operate on all the memory blocks in each iteration? Then we can remove the `if-else` and make the benchmark focus more on the memory block? As a comparison, we can create a byte array, a long array, an offheap array, and also operate on them in each iteration.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org