Intel® Itanium® Processor Family SAL Specification
This document describes the functionality of the System Abstraction Layer (SAL) for Itanium® architecture-based systems.
This document specifies requirements to develop platform firmware for Itanium architecture-based systems. A companion document, the Unified Extensible Firmware Interface Specification, describes additional interfaces that must be implemented to access devices on the platform. The Unified Extensible Firmware Interface Specification is a requirement for Itanium architecture-based firmware.
This document is intended for firmware designers, system designers, and writers of diagnostic and low level operating system software. This document is an architectural specification describing the platform-dependent firmware interfaces needed to support the objectives listed below. It does not require a specific implementation, nor is it intended to document PC infrastructure specifications.
The primary objectives of Itanium architecture-based firmware are to:
• Enable boot of Itanium architecture-based operating systems.
• Ensure that the firmware interfaces encapsulate the platform implementation differences within the hardware abstraction layers and device driver layers of operating systems.
• Separate the platform abstraction from the processor abstraction.<BR>• Enable platform differentiation, hardware innovation, and optimization of Itanium architecture-based platforms.
• Support the scaling of systems from the low-end to the high-end including servers, workstations, mainframe alternatives, and supercomputers. Features supported will include high availability, error logging and recovery, large memory support, multiprocessing, and broader and deeper I/O hierarchies (possibly greater than 100 I/O cards).
• While using Itanium instructions is preferred, IA-32 BIOS code can be used in SAL. The extent of the IA-32 BIOS reuse is implementation-dependent, but all SAL entry points from the Processor Abstraction Layer (PAL) will use the Itanium system environment.
Texas Advanced Computing Center demonstrates OSPRay* ray tracing for modelling groundwater rock structure.
Enable clients to access high performance computing with the Intel® Cluster Ready program.
Intel® Xeon® processor E5-2600 v3 product family improves genome mapping workflows for personalized medicine.
DownUnder GeoSolutions can process data in minutes instead of hours with the Intel® Xeon Phi™ coprocessor.
University of Tennessee uses Intel® Xeon Phi™ coprocessor to extract insights with desktop PCs.
See how Intel is addressing tomorrow's HPC needs for digital content, weather, financial services, and more.