Overview

In March 2004, at the Intel Developer Forum in San Francisco, company CEO Craig Barrett confirmed the existence of a technology that, while unacknowledged until then, was a widely known "open secret:" Intel would soon ship 32-bit processors endowed with 64-bit extensions. This statement opened a new chapter in the history of desktop computing one that begins a long-term migration away from 32-bit processors to the new, extraordinary capacities of 64-bit architectures.

The 64-bit extensions referred to in that speech were initially dubbed IA-64e by Intel, which stood for Intel Architecture 64-bit extensions. They were ultimately renamed Intel Extended Memory 64 Technology or Intel EM64T, which is how this book refers to them.

The reason the technology uses the term extensions is that the technology is embedded in the 32-bit Intel Architecture core (IA-32) that has seen service in several releases of Pentium 4 and Intel Xeon processors. Intel EM64T provides extensions to that core, which enable the 64-bit operations, so applications can access larger amounts of memory. Many features of the IA-32 processor core remain the same or almost the same, as I explain later in greater detail.

The key feature of Intel EM64T, and of 64-bit computing in general, is a much larger address space, that is, access to very significantly more memory. In theory, 32-bit software is limited to 4 billion possible addresses, meaning that a 32-bit processor...