At the end of the day, all transactional RDBMSs need to do the same things, and if they're mature and optimized enough, they'll all reach the same physical limits of what the underlying disk can handle. There will be small differences, but all numbers will be in the same magnitude.
Switching from MySQL to PostGRES in the above scenario wouldn't do much for performance, some operational tasks might change, complex queries might change in speed, but the baseline of simple updates/selects per second won't really change.
To get magnitudes more performance, you need to use a different model. Redis is a NoSQL key-value-store which keeps the entire dataset in memory and occasionally flushes changes to disk. Of course that's going to be faster than a system which flushes all changes to disk individually.
Switching from MySQL to PostGRES in the above scenario wouldn't do much for performance, some operational tasks might change, complex queries might change in speed, but the baseline of simple updates/selects per second won't really change.
To get magnitudes more performance, you need to use a different model. Redis is a NoSQL key-value-store which keeps the entire dataset in memory and occasionally flushes changes to disk. Of course that's going to be faster than a system which flushes all changes to disk individually.