In further bad news for customers of Intel's Skylake chips, PC World is reporting that Intel has is issuing a microcode update which removes a feature present at shipping which allowed safely and easily overclocking the chips by increasing their base clock frequency (archived). Normally overclocking is done safely on other chips through increasing the multiplier applied to the base clock. This is because in other chips the base clock frequency is used to derive the frequency of a number of buses on the system which can begin exhibiting aberrant behavior when the base clock frequency is molested. Skylake however divorced the base clock from other system buses which reintroduced the possibility of overclocking the chips by increasing the base clock while leaving the locked multiplier alone.
Until Intel announced they would be using a microcode update to remove this feature it prominently allowed any chip in the Skylake family to be safely overclocked, though only premium priced K series parts could be overclocked using the legacy method of increasing the multiplier. This episode is just one in a long series of abuses issued to customers by Intel. Customers of Skylake chips are being forced by Intel to make a choice between keeping this feature or get microcode updates which fix bugs that prevented the chips from properly doing math1 at the time they shipped.
It seems Intel is getting a head start on their long term roadmap to increase energy efficiency at the price of computing performance.
The original microcode on a number of Skylake chips lead the chips to freeze during mathematical benchmarks that involved searching for Mersenne prime numbers. ↩