1Introduction
1.1Overview
The Itanium 2 processor employs Explicitly Parallel Instruction Computing (EPIC) design concepts for a tighter coupling between hardware and software. In this design style, the interface between hardware and software is designed to enable the software to exploit all available compile- time information, and efficiently deliver this information to the hardware. It addresses several fundamental performance bottlenecks in modern computers, such as memory latency, memory address disambiguation, and control flow dependencies. The EPIC constructs provide powerful architectural semantics, and enable the software to make global optimizations across a large scheduling scope, thereby exposing available Instruction Level Parallelism (ILP) to the hardware. The hardware takes advantage of this enhanced ILP, and provides abundant execution resources. Additionally, it focuses on dynamic run-time optimizations to enable the compiled code schedule to flow at high throughput. This strategy increases the synergy between hardware and software, and leads to greater overall performance.
The Itanium 2 processor provides a 6-wide and 8-stage deep pipeline, running at up to 1.66 GHz. This provides a combination of abundant resources to exploit ILP as well as increased frequency for minimizing the latency of each instruction. The resources consist of six integer units, six multimedia units, two load and two store units, three branch units, two extended-precision floating point units, and one additional single-precision floating point unit. The hardware employs dynamic prefetch, branch prediction, a register scoreboard, and non-blocking caches to optimize for compile-time non-determinism. Three levels of on-die cache minimize overall memory latency. This includes up to a 9 MB L3 cache, accessed at core speed, providing up to 84.8 Gb/sec of data bandwidth. The system bus is designed to support up to four processors (on a single system bus), and can be used as an effective building block for very large systems. The balanced core and memory subsystem provide high performance for a wide range of applications ranging from commercial workloads to high performance technical computing.
The Itanium 2 processor supports a range of computing needs and configurations from a 2-way to large SMP servers. This document provides the electrical, mechanical and thermal specifications for the Itanium 2 processor for use while using systems with Itanium 2 processors.
1.2Processor Abstraction Layer
The Itanium 2 processor requires implementation-specific Processor Abstraction Layer (PAL) firmware. PAL firmware supports processor initialization, error recovery, and other functionality. It provides a consistent interface to system firmware and operating systems across processor hardware implementations. The Intel® Itanium® Architecture Software Developer’s Manual, Volume 2: System Architecture, describes PAL. Platforms must provide access to the firmware address space and PAL at reset to allow Itanium 2 processors to initialize.
The System Abstraction Layer (SAL) firmware contains platform-specific firmware to initialize the platform, boot to an operating system, and provide runtime functionality. Further information about SAL is available in the Intel® Itanium® Processor Family System Abstraction Layer Specification.
