We've created the first vectorized Quicksort

Google has open-sourced a vectorized Quicksort algorithm, which uses SIMD instructions to improve performance on modern CPUs. The new Quicksort implementation uses AVX-512 instructions to process multiple data elements in parallel, resulting in a significant speedup on systems that support it. The code is designed to be performance portable, meaning it can automatically switch between using SIMD instructions and scalar code based on the capabilities of the system it's running on. This allows it to achieve optimal performance on both modern and older systems. The implementation also includes a number of optimizations, including the use of a hand-tuned recursive function and a hybrid approach that uses a combination of recursive and iterative sorting.