To be clear, your program ran at half speed, right? That's far worse than doubling the time spent in memory-management functions.

Yes, our program ran at half speed.

It is quite bad if changing the allocator made your program run at half speed. It doesn't matter if the library itself consumed more CPU or caused more wait (guessing that's the case here) due to concurrency or excessive syscalls.

The developer helped us a lot to improve the speed and now it is almost at the same level as jemalloc for us