high concurrency scenario

Scheduling too many tasks at the same time will result in the loss of some tasks, about 1%.
Individual tasks are small and fast. The average single execution time is about 10 seconds, with about 20+ task commits per second.
Is this a platform exception? How to optimize it?
Server resource utilization is less than 50% and there are no resource issues.