欢迎访问ic37.com |
会员登录 免费注册
发布采购

80960MC 参数 Datasheet PDF下载

80960MC图片预览
型号: 80960MC
PDF下载: 下载PDF文件 查看货源
内容描述: 具有集成浮点单元和内存管理单元采用嵌入式32位微处理器 [EMBEDDED 32-BIT MICROPROCESSOR WITH INTEGRATED FLOATING-POINT UNIT AND MEMORY MANAGEMENT UNIT]
分类和应用: 内存管理单元微处理器
文件页数/大小: 39 页 / 401 K
品牌: INTEL [ INTEL ]
 浏览型号80960MC的Datasheet PDF文件第6页浏览型号80960MC的Datasheet PDF文件第7页浏览型号80960MC的Datasheet PDF文件第8页浏览型号80960MC的Datasheet PDF文件第9页浏览型号80960MC的Datasheet PDF文件第11页浏览型号80960MC的Datasheet PDF文件第12页浏览型号80960MC的Datasheet PDF文件第13页浏览型号80960MC的Datasheet PDF文件第14页  
80960MC  
REGISTER  
CACHE  
LOCAL REGISTER SET  
ONE OF FOUR  
LOCAL  
REGISTER SETS  
R
0
R
15  
0
31  
Figure 3. Multiple Register Sets Are Stored On-Chip  
1.1.7 Memory Management and Protection  
1.1.8 Floating-Point Arithmetic  
The 80960MC is ideal for multitasking applications  
that require software protection and a large address  
space. To ensure the highest level of performance  
possible, the memory management unit (MMU) and  
translation look-aside buffer (TLB) are contained on-  
chip.  
In the 80960MC, floating-point arithmetic is an  
integral part of the architecture. Having the floating-  
point unit integrated on-chip provides two advan-  
tages. First, it improves the performance of the chip  
for floating-point applications, since no additional  
bus overhead is associated with floating-point calcu-  
lations, thereby leaving more time for other bus oper-  
ations such as I/O. Second, the cost of using  
The 80960MC supports a conventional form of  
demand-paged virtual memory in which the address  
space is divided into 4-Kbyte pages. Studies indicate  
that a 4-Kbyte page is the optimum size for a broad  
range of applications.  
floating-point operations is reduced because  
separate coprocessor chip is not required.  
a
The 80960MC floating-point (real-number) data  
types include single-precision (32-bit), double-preci-  
sion (64-bit) and extended precision (80-bit) floating-  
point numbers. Any registers may be used to  
execute floating-point operations.  
Each page table entry includes a 2-bit page rights  
field that specifies whether the page is a no-access,  
read-only, or read-write page. This field is inter-  
preted differently depending on whether the current  
task (process) is executing in user or supervisor  
mode, as shown below:  
The processor provides hardware support for both  
mandatory and recommended portions of IEEE  
Standard 754 for floating-point arithmetic, including  
all arithmetic, exponential, logarithmic and other  
transcendental functions. Table 3 shows execution  
times for some representative instructions.  
Rights  
00  
01  
10  
11  
User  
Supervisor  
Read-Only  
Read-Write  
Read-Write  
Read-Write  
No Access  
No Access  
Read-Only  
Read-Write  
6
PRELIMINARY