GTX4060 Laptop GPU - Stable diffusion 1024x1024 at 20 steps for SDXL only needs 30 seconds. If you are using Flux, then it will be 3minutes. So maye you want to check whether you updated cuda and pytorch to the latest version. If you use the money to buy a Macbook Pro, you could have gotten a GTX4080 and it will definitely be faster than Macbook pro by a mile.