Real-world latency impact of turning on all the IPS signatures. Is it crippling?

Hey everyone, new to the team and the NGFW world. We're deploying a new Palo Alto and the security team wants us to enable the entire IPS signature database for maximum protection.

Our senior engineer warned that enabling *all* signatures can sometimes introduce a lot of latency, especially for our internal app servers that talk to each other. The datasheet numbers are one thing, but I'm trying to understand what happens in practice.

Has anyone run into this? Did you see a noticeable hit on east-west traffic or user VPN performance? We're especially worried about our SQL traffic between application tiers. Any guidance on a more surgical approach would be a lifesaver.

Thanks in advance!