I'm evaluating the STM32U585 TLS performance using the eval board B-U585I-IOT02A. The example application code for MQTT using TLS did not include code on how to load a client key for client authentication but I was able to add that in. The problem I see now is that when using private key the MCU takes 13 seconds to do the RSA computations during the handshake with the server to setup the connection. I track the delay to the following function _nx_crypto_huge_number_mont_power_modulus
Why does this take so long? Does the MCU have hardware acceleration that handles this and is just not being used? Is there something else that I'm doing wrong?
