Keeping Technical Professionals Up-To-Date  

 
 Home
 
 Courses

    PC Firmware
    (UEFI, Tiano and BIOS)

    PC Architecture

    CPU Architecture

    Linux

 

 
Schedule a Course
     Events in Asia
    
 
 About Techstream
 
 Contact Us
 
 

 

Course 205 Click here to download a printable PDF file of this page
 

Itanium® System Architecture Overview
 A 2-day Course

With its massive on-chip resources, several parallel execution pipes, and support for large, fast caches the Itanium Architecture creates new opportunities and challenges for both hardware, system-level, and application-level developers. This course will bring you up-to-speed on Itanium’s architectural features at both the processor and system levels as well as their impact on software development approaches.

You will benefit from this workshop if you

  • Are evaluating Intel’s Itanium processors for future designs

  • Want to keep up-to-date with the latest features and capabilities in the Itanium processor family

You will learn

  • What traditional bottlenecks the Itanium Architecture eliminates

  • The innovative register structure of the Itanium family CPUs

  • How to exploit the CPU's capabilities when developing system or application software

  • The key features of the chipsets that support Itanium-class processors

Prerequisites

A technical background is expected, including a solid understanding of general computer architecture and terminology. Familiarity with the IA-32 hardware architecture at the server or workstation level is helpful but not required.

The training approach

  • Up to date information: We update the materials before every event.

  • Straightforward explanations: Technical concepts and terms are explained in English. You will walk away with a thorough understanding of what the Itanium Architecture brings to the table and how to exploit its capabilities.

Workshop topics

Itanium’s solution to traditional CPU architecture bottlenecks

  • EPIC: Explicitly Parallel Instruction Computing – No more sequential semantics

  • Remove branches with predicated instructions

  • Minimize impact of memory latency

  • Decrease procedure call overhead

  • Minimize loop optimization overhead

  • Massive memory resources

  • Massive register resources

  • The Register Stack Engine (RSE)

The Itanium micro architecture

  • Processor pipelines

  • Processor block diagram

  • Instruction processing

  • Dispersal logic

  • Execution units

  • Pipeline control

  • Exception handler

  • Cache subsystem

    • Level 1 instruction and data caches

    • Level 2 and Level 3 cache

    • Cache coherency

    • ALAT and TLB

  • IA-32 execution

  • Memory addressing

    • System memory model

    • Virtual memory model

    • Memory alignment

  • I/O addressing

Itanium Product Family (IPF) chipsets

  • Itanium Tukwila - Socket- and chipset-compatible with Xeon

    • Integrated memory controller

    • QuickPath replaces the Front Side Bus

      • Intel standardized Nortbridges and Southbridges

  • The E8870 chipset

    • Block diagram

    • Chipset highlights

      • Scalable Node Controller (SNC)

      • DDR Memory Hub (DMH)

      • Scalability Port Switch (SPS)

      • Server I/O Hub (SIOH)

      • PCI/PCI-X Hub (P64H2)

      • Interface Control Hub (ICH)

      • Firmware Hub (FWH)

A new way to boot

  • Why change?

  • The Itanium boot process

    • Tiano - The Intel Platform Innovation Framework for EFI

    • EFI – Extensible Firmware Interface

    • The operating system loader

  • Runtime services & protocols

  • Firmware address space

  • Firmware Interface Table (FIT)

  • New disk layout

Tools for integration and test

  • Test Access Port – TAP

    • Interface

    • TAP registers

    • Instructions for Itanium TAP

    • Reset behavior

  • Integration tools

    • In-target probe

    • Logical modeling tools

Itanium processor package

  • Cartridge features

  • Mechanical dimensions

Application level software

  • System environment

    • Instruction set transitions

    • Compiler to processor communication

    • Control and data speculation

    • Predication

    • Branching

    • Register rotation

    • Floating-point architecture

  • Execution environment

    • General registers

    • Floating point registers

    • Predicate registers

    • Branch registers

    • Application registers

    • User mask

    • Processor identification

  • Application programming model

    • Integer computation

    • Compare instructions and predication

    • Memory address instructions

    • Data pre-fetch hints

    • Branch instructions

System level software

  • Itanium system environment

    • System environment

    • Privilege levels

    • System state registers

    • Virtual memory support

    • Region registers

    • Translation Lookaside Buffer (TLB)

  • Interruptions

    • Four types of defined interruptions

    • Interruption programming model

    • Interruption handling

  • Register Stack Engine (RSE) and backing store

  • Debugging and performance monitoring

  • Context management

Available operating systems

  • Unix

  • Linux

  • Windows

  • VMS

  • Others

 

Courses  Schedule a Course  |  About Us  |  Contact Us  |  Home

 

Copyright © 2005-2013, Techstream Inc. All rights reserved