PA-RISC
http://baike.baidu.com/view/167703.htm
PA-RISC,由惠普公司开发的一种处理器指令集架构,属于精简指令集架构。PA 是指精准指令集架构(Precision Architecture),所以它又被称为惠普精准指令集架构(Hewlett Packard Precision Architecture,HP/PA)。它首次出现于1986年2月26日,HP 3000 930系列以及HP 9000 840模式处理器之中。
它之后被惠普公司与英特尔联合开发的Itanium指令集架构所取代。
http://en.wikipedia.org/wiki/PA-RISC
PA-RISC
Designer | Hewlett-Packard |
---|---|
Bits | 64-bit (32→64) |
Introduced | 1986 (1996 PA-RISC 2.0) |
Version | 2.0 (1996) |
Design | RISC |
Encoding | Fixed |
Branching | Compare and branch |
Endianness | Big |
Extensions | Multimedia Acceleration eXtensions (MAX), MAX-2 |
Open | No |
Registers | |
General purpose | 32 |
Floating point | 32 64-bit (16 64-bit in PA-RISC 1.0) |
HP PA-RISC 7300LC Microprocessor
PA-RISC is an instruction set architecture (ISA) developed by Hewlett-Packard. As the name implies, it is a reduced instruction set computer (RISC) architecture, where the PA stands for Precision Architecture. The design is also referred to as HP/PA for Hewlett Packard Precision Architecture.
The architecture was introduced on 26 February 1986 when the HP 3000 Series 930 and HP 9000 Model 840 computers were launched featuring the first implementation, the TS1.[1][2]
PA-RISC has been succeeded by the Itanium (originally IA-64) ISA jointly developed by HP and Intel.[3] HP stopped selling PA-RISC-based HP 9000 systems at the end of 2008 but will support servers running PA-RISC chips until 2013.[4]
History[edit]
In the late 1980s HP was building four series of computers, all based on CISC CPUs. One line was the IBM PC compatible Intel i286 based Vectra Series started 1986. All others were non-Intel systems. One of them was the HP Series 300 of Motorola 68000-based workstations, another Series 200 line of technical workstations based on a custom silicon on sapphire (SOS) chip design, the SOS based 16-bit HP 3000 classic series and finally the HP 9000 Series 500 minicomputers, based on their own (16 and 32-bit) FOCUS microprocessor. HP planned to use PA-RISC to move all of their non-PC compatible machines to a single RISC CPU family.
Precision Architecture was introduced in 1986. It had thirty-two 32-bit integer registers and sixteen 64-bit floating-point registers. The number of floating-point registers was doubled in the 1.1 version to 32 once it became apparent that 16 were inadequate and restricted performance. The architects included Allen Baum, Hans Jeans, Michael J. Mahon, Ruby Bei-Loh Lee, Russel Kao, Steve Muchnick, Terrence C. Miller, David Fotland, and William S. Worley.[5]
The first implementation was the TS1, a central processing unit built from discrete transistor-transistor logic (74F TTL) devices. Later implementations were multi-chip VLSI designs fabricated in NMOS processes (NS1 and NS2) and CMOS (CS1 and PCX). They were first used in a new series of HP 3000 machines in the late 1980s – the 930 and 950, commonly known at the time as Spectrum systems, the name given to them in the development labs. These machines ran MPE/iX. The HP 9000 machines were soon upgraded with the PA-RISC processor as well, running the HP-UX version ofUNIX.
Other operating systems ported to the PA-RISC architecture include Linux, OpenBSD, NetBSD and NEXTSTEP.
An interesting aspect of the PA-RISC line is that most of its generations have no Level 2 cache. Instead large Level 1 caches are used, formerly as separate chips connected by a bus, and now integrated on-chip. Only the PA-7100LC and PA-7300LC had L2 caches. Another innovation of the PA-RISC was the addition of vectorized instructions (SIMD) in the form of MAX, which were first introduced on the PA-7100LC.
The ISA was extended in 1996 to 64-bits, with this revision named PA-RISC 2.0. PA-RISC 2.0 also added fused multiply–add instructions, which help certain floating-point intensive algorithms, and the MAX-2 SIMD extension, which provides instructions for accelerating multimedia applications. The first PA-RISC 2.0 implementation was the PA-8000, which was introduced in January 1996.
CPU specifications[edit]
Model | Marketing name | Year | Frequency [MHz] | Memory Bus [MB/s] | Process [µm] | Transistors [millions] | Die size [mm²] | Power [W] | Dcache [kB] | Icache [kB] | L2 cache [MB] | ISA | Notes |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
TS-1 | ? | 1986 | 8 | ? | ? | — | — | ? | ? | ? | — | 1.0 | |
CS-1 | ? | 1987 | 8 | ? | 1.6 | 0.164 | 72.93 | 1 | — | 0.25 | — | 1.0 | [6] |
NS-1 | ? | 1987 | 25/30 | ? | 1.5 | 0.144 | 70.56 | ? | ? | ? | — | 1.0 | [7] |
NS-2 | ? | 1989 | 27.5/30 | ? | 1.5 | 0.183 | 196 | 27 | 512 | 512 | — | 1.0 | [8] |
PCX | ? | 1990 | ? | ? | ? | ? | ? | ? | ? | ? | ? | 1.0 | |
PCX-S | PA-7000 | 1991 | 66 | ? | 1.0 | 0.58 | 201.6 | ? | 256 | 256 | — | 1.1a | |
PCX-T | PA-7100 | 1992 | 33–100 | ? | 0.8 | 0.85 | 196 | ? | 2048 | 1024 | — | 1.1b | |
PCX-T | PA-7150 | 1994 | 125 | ? | 0.8 | 0.85 | 196 | ? | 2048 | 1024 | — | 1.1b | |
PCX-T' | PA-7200 | 1994 | 120 | 960 | 0.55 | 1.26 | 210 | 30 | 1024 | 2048 | — | 1.1c | |
PCX-L | PA-7100LC | 1994 | 60–100 | ? | 0.75 | 0.9 | 201.6 | 7–11 | — | 1 | 2 | 1.1d | |
PCX-L2 | PA-7300LC | 1996 | 132–180 | ? | 0.5 | 9.2 | 260.1 | ? | 64 | 64 | 0–8 | 1.1e | |
PCX-U | PA-8000 | 1996 | 160–180 | 960 | 0.5 | 3.8 | 337.68 | ? | 1024 | 1024 | — | 2.0 | |
PCX-U+ | PA-8200 | 1997 | 200–240 | 960 | 0.5 | 3.8 | 337.68 | ? | 2048 | 2048 | — | 2.0 | |
PCX-W | PA-8500 | 1998 | 300–440 | 1920 | 0.25 | 140 | 467 | ? | 1024 | 512 | — | 2.0 | [9] |
PCX-W+ | PA-8600 | 2000 | 360–550 | 1920 | 0.25 | 140 | 467 | ? | 1024 | 512 | — | 2.0 | [9] |
PCX-W2 | PA-8700(+) | 2001 | 625–875 | 1920 | 0.18 | 186 | 304 | <7.1@1.5 V | 1536 | 768 | — | 2.0 | |
Mako | PA-8800 | 2003 | 800–1000 | 6400 | 0.13 | 300 | 361 | ? | 768/core | 768/core | 0 or 32 | 2.0 | |
Shortfin | PA-8900 | 2005 | 800–1100 | 6400 | 0.13 | ? | ? | ? | 768/core | 768/core | 64 | 2.0 |
|
PA-RISC的更多相关文章
- Delphi : Analyze PE file headers?
Analyze PE file headers? { You'll need a OpenDialog to open a Exe-File and a Memo to show the file i ...
- [Linux] 001 预备知识
Unix 1965年 MIT,通用电气(GE),AT&T 的贝尔实验室联合开发 项目名称:Multics 目标:开发一种交互式的,具有多道程序处理能力的分时操作系统 后来:贝尔实验室宣布退出 ...
- ARM概论(Advanced RISC Machines)
简介 ARM7是32 位通用微处理器ARM(Advanced RISC Machines)家族中的一员,具有比较低的电源消耗和良好的性价比, 基于(精简指令)RISC结构,指令集和相关的译码机制与微程 ...
- RISC指令集的五个周期
RISC指令集的五个周期 RISC(reduced instruction set computer,精简指令集计算机)简称为精简指令集.RISC把执行指令的精力主要放在了经常使用的指令上面.本文主要 ...
- 2.1 CMMI2级——7个PA简述
摘要: 阶段式的CMMI没有1级,最开始的级别就是2级.一个处于“无序化”生产的软件公司,要进行过程改进,首要是改进什么呢?2级告诉你,我们需要从计划.计划跟踪.需求管理.采购.度量.配置管理.质量保 ...
- [RouterOS] ROS对接碧海威或PA等流控实现完美流控详细教程(附脚本全免费)
前言: 经常在群里看到不少朋友争论海蜘蛛 ROS 维盟 爱快 碧海威 流控大师 Woyos等等软路由,哪个好.实际上,网络产品是复杂的,现在的软路由功能上已经远远不是单独的路由了.每种产品都有他本身的 ...
- 小数量宽带用户的福音,Panabit 云计费easyradius 接口隆重发布,PA宽带计费系统
PA接口在早前就发布了,但是一直迟迟没有发布官方说明文档,由于最近问的客户较多,特写了这篇文档 由于PA使用标准radius认证协议,所以用户需要在本地搭建一个计费,由于大部分用户的数量只有几百个,不 ...
- LNA和PA
低噪声放大器(Low Noise Amplifier) -------------LNA 功率放大器(Power Amplifier)---------------------PA LNA是低噪声放大 ...
- risc与cisc
RISC(精简指令集计算机)和CISC(复杂指令集计算机)是当前CPU的两种架构.它们的区别在于不同的CPU设计理念和方法. 早期的CPU全部是CISC架构,它的设计目的是要用最少的机器语言指令来完成 ...
- 有两个指针pa,pb分别指向有两个数,a,b,请写一个函数交换两个指针的指向,也就是让pa指向b,让pb指向a
题目:有两个指针pa,pb分别指向有两个数,a,b,请写一个函数交换两个指针的指向,也就是让pa指向b,让pb指向a,具体实现如下: #include<stdlib.h> #include ...
随机推荐
- STL之set容器的总结
最近做了很多题型,都是用简单的STL就解决了,深刻的感觉到STL的伟大力量,但是本人在遇到问题的时候还是喜欢用常规的算法去解决问题,脑袋笨没办法,有时候根本想不到用STL去解决一些问题 往往都是砍了网 ...
- 计算机、程序和java概述
1.1 计算机.程序和java概述计算机包括硬件(hardware)软件(software)两部分.硬件包括计算机中看得见的物理部分,软件提供看不见的指令,指令控制硬件并且使得硬件完成特定的任务.一台 ...
- BZOJ3507 [Cqoi2014]通配符匹配 【哈希 + 贪心】
题目 几乎所有操作系统的命令行界面(CLI)中都支持文件名的通配符匹配以方便用户.最常见的通配符有两个,一个 是星号(""'),可以匹配0个及以上的任意字符:另一个是问号(&quo ...
- nginx 变量 + lua
nginx变量使用方法详解(8) nil.null与ngx.null 发现一个nginx LUA开发Web App的框架 nginx是个好东西, nginx的openrtsy发行版本更是个好东西. 今 ...
- UVa1476 Error Curves
画出函数图像后,发现是一个类似V字型的图. 可以用三分法找图像最低点 WA了一串,之后发现是读入优化迷之被卡. /*by SilverN*/ #include<iostream> #inc ...
- zoj 2760(网络流+floyed)
How Many Shortest Path Time Limit: 10 Seconds Memory Limit: 32768 KB Given a weighted directed ...
- Python Challenge 第一关
偶然在网上看到这个,PYTHON CHALLENGE,利用Python语言闯关,觉得挺有意思,就记录一下. 第0关应该算个入口吧,试了好几次才试出来,没什么代码就不写了.计算一个结果出来就行. 第一关 ...
- 用jQuery File Upload实现简单的文件上传
FORM中的代码: {# file_path #} <div class="form-group"> <label class="control-lab ...
- Gaugecontrol(测量仪器图形控件)
digital 数字类 circularfull 整圆 circularhalf 半圆 circularquarter 四分之一圆 circularThreeFourth 四分之三圆 linear h ...
- 洛谷——P2296 寻找道路
P2296 寻找道路 题目描述 在有向图G 中,每条边的长度均为1 ,现给定起点和终点,请你在图中找一条从起点到终点的路径,该路径满足以下条件: 1 .路径上的所有点的出边所指向的点都直接或间接与终点 ...