UEFI replaced BIOS after Microsoft made it a Windows 8 requirement

BIOS was designed for 16-bit PCs and by the 1990s it was the bottleneck on larger machines. Intel started the effort that became EFI in 1998 — initially the Intel Boot Initiative — for Itanium servers where BIOS limitations had become too restrictive, and shipped EFI on its Itanium workstations from 2000.

In 2005 Intel contributed EFI 1.10 to the Unified EFI Forum, which developed it as UEFI — a specification owned by an industry consortium rather than one company, with version 2.0 released in January 2006. The technology worked; most consumer PCs still booted through BIOS compatibility mode.

What changed everything was a certification requirement. Microsoft's Windows 8 client logo rules made UEFI firmware mode a requirement, and Microsoft's own documentation is explicit: most consumer computers had continued to boot using BIOS firmware, and that changed with Windows 8, when UEFI firmware mode became a Windows 8 client logo requirement — bringing GPT partitioning, boot from >2.2 TB drives, and Secure Boot.

Hand firmware to an industry forum; certify its use

An industry forum's spec freed firmware from one vendor's control.

GPT and >2.2 TB boot support solved problems BIOS physically could not.

Secure Boot gave OEMs a security story Windows 8 certification demanded.

The logo requirement forced every PC maker to ship UEFI at once.

A standard's technology can be ready for years; adoption arrives when a platform owner makes it a certification requirement — the requirement, not the spec, moves the market.

After Windows 8, UEFI became the firmware of every consumer PC, and the same certification carried GPT and Secure Boot into the mainstream; the UEFI Forum continues to own and evolve the specification that replaced BIOS.