Im creating a game + engine and speaking from personal experience/my use case, hyperthreading was less performant than (praying to the CPU thread allocation god

Im creating a game + engine and speaking from personal experience/my use case, hyperthreading was less performant than (praying to the CPU thread allocation god) each thread utilizing its own core. I decided to max out the number of threads by using std::thread::hardware_concurrency() / 2 - 1. (i.e. number of cores - 1 ). I'm working with a std::vector