A Kernel Maintainer's Prediction On The CPU Architecture Landscape For 2030

In addition to talking about code/hardware obsolescence from the Linux kernel, prominent upstream Linux kernel developer Arnd Bermann also presented at last week's Linux Plumbers Conference on the current SoC landscape and sharing his predictions for ten years down the road.

Bergmann gave a talk in addition to the obsolescence one on the "SoC support lifecycle in the kernel" when talking about changes in SoC/CPU architectures and how that has evolved over time and the Linux kernel embracing the changes while also identifying code that has reached the end of its useful life.

One of the interesting parts of his presentation at the end is on his predictions by the year 2030. The longtime Linux developer who is currently employed by Linaro envisions x86-64, ARMv8+, and RISC-V splitting the CPU architecture market.... Not too surprising aside from perhaps not having much confidence in POWER over the next decade. He also thinks over the next decade more SoC vendors will better engage with upstream Linux on their driver upstreaming, ARMv7 chips will still be shipping while the last models released, more creative multi-chip packages like chipets, IBM Z still being profitable, new ISAs will prove to be unsuccessful, the start of 128-bit computing, and everything else will be obsolete.

Find more of Arnd's thoughts and his personal predictions via his presentation embedded above and also the PDF slide deck.