I am preparing a school presentation about underclocking/overclocking and have read several articles about the CPU multiplier during my research. Some of the articles have claimed that the CPU itself adjusts the multiplier, others have written that the BIOS firmware or the OS adjusts it. I have not saved the link to the last statement, but most of the answers here in this forum and others seem to imply that the CPU itself is responsible for adjusting the multiplier/ frequency.
In my understanding, the maximum CPU multiplier is determined by the CPU's maximum performance. A CPU that has 3,5 Gigahertz combined with a system that runs at 100 Megahertz base clock speed has a maximum multiplier of 35. The motherboard/BIOS or the OS can directly control the multiplier by sending instructions to certain registers within the CPU so it runs fewer instructions per second. Is this assumption correct?