HP ProLiant DL590/64 ProLiant DL590/64 Server Technology - Page 3

ProLiant DL590/64 Server Technology, 64-Bit Computing Overview, Early 32-bit Intel Architecture IA-32

Get HP ProLiant DL590/64 PDF manuals and user guides View all HP ProLiant DL590/64 manuals
Add to My Manuals
Save this manual to your list of manuals

Page 3 highlights

TC020702TB TECHNOLOGY BRIEF (cont.) ... 64-BIT COMPUTING OVERVIEW At its simplest level, 64-bit processing means the registers and execution units within the processor can access addresses and manipulate data that are 64 bits wide. Although 32-bit microprocessor architecture has 64-bit data paths, its internal registers and execution units only access addresses and manipulate data that are 32 bits wide. True 64-bit computing requires not only the 64-bit characteristics of the processor, but also a 64-bit operating system (OS). The characteristics of a 64-bit computing environment include very large files and file systems, large amounts of system and virtual memory, and a large memory addressing space. For most enterprise applications, the greatest immediate impact of 64-bit computing is the ability to address a significantly larger pool of system memory-as much as 64-gigabytes (GB) with the first 64-bit processor and chipset from Intel and, theoretically, more than 700 terabytes with subsequent processors on the Intel roadmap. Addressing larger system memory pools will allow applications to run faster while handling larger data sets. Early 32-bit Intel Architecture (IA-32) processors could directly address only 4 GB of system memory; more recent IA-32 processors can directly address up to 16 GB of system memory. Initial implementations of 64-bit Intel Architectures (IA-64) will allow direct, high-speed access to 64 GB of system memory. The large 64-bit address space allows 4 billion times as many memory addresses as 32-bit systems, resulting in a theoretical physical memory limit into the exabytes. As memory density increases and relative cost of memory decreases, it becomes feasible to implement system memory in the hundreds of GB, enabling significant performance improvements for large, memory-intensive enterprise applications. The benefits of larger system memory pools include: • Faster access to huge amounts of data in memory-a significant advantage for customers who use large databases • Increased capacity for more users • Greater application performance due to less page swapping and fewer page faults • Easier communication between programs to share code and data • Better manipulation of huge data sets • Solutions for larger and more complex requirements Customers with power-hungry applications such as decision support, database management, data mining, data warehousing, and e-commerce will see significant performance improvements through the adoption of 64-bit solutions. Applications requiring higher floating point capabilities, such as number crunching, analytics, rendering, and high-performance technical computing (HPTC) will also see performance advantages over 32-bit systems. As compilers mature, many software vendors and HPTC companies will be investigating IA-64 systems as an alternative to UNIX platforms. Where the benefits of IA-64 can be recognized, many software vendors and companies with custom code are using IA-64 systems as a development platform for porting their code. Since any code that is ported on IA-64 systems can easily be migrated to the follow-on platform based on the next-generation 64-bit processor, it makes sense that these companies start early to increase their competitiveness in the market. 3

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
T
ECHNOLOGY
B
RIEF
(cont.)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
64-B
IT
C
OMPUTING
O
VERVIEW
At its simplest level, 64-bit processing means the registers and execution units within the processor
can access addresses and manipulate data that are 64 bits wide.
Although 32-bit microprocessor
architecture has 64-bit data paths, its internal registers and execution units only access addresses
and manipulate data that are 32 bits wide.
True 64-bit computing requires not only the 64-bit
characteristics of the processor, but also a 64-bit operating system (OS).
The characteristics of a
64-bit computing environment include very large files and file systems, large amounts of system
and virtual memory, and a large memory addressing space.
For most enterprise applications, the greatest immediate impact of 64-bit computing is the ability
to address a significantly larger pool of system memory—as much as 64-gigabytes (GB) with the
first 64-bit processor and chipset from Intel and, theoretically, more than 700 terabytes with
subsequent processors on the Intel roadmap.
Addressing larger system memory pools will allow
applications to run faster while handling larger data sets.
Early 32-bit Intel Architecture (IA-32)
processors could directly address only 4 GB of system memory; more recent IA-32 processors can
directly address up to 16 GB of system memory.
Initial implementations of 64-bit Intel
Architectures (IA-64) will allow direct, high-speed access to 64 GB of system memory. The large
64-bit address space allows 4 billion times as many memory addresses as 32-bit systems, resulting
in a theoretical physical memory limit into the exabytes.
As memory density increases and relative cost of memory decreases, it becomes feasible to
implement system memory in the hundreds of GB, enabling significant performance improvements
for large, memory-intensive enterprise applications.
The benefits of larger system memory pools
include:
•
Faster access to huge amounts of data in memory—a significant advantage for customers who
use large databases
•
Increased capacity for more users
•
Greater application performance due to less page swapping and fewer page faults
•
Easier communication between programs to share code and data
•
Better manipulation of huge data sets
•
Solutions for larger and more complex requirements
Customers with power-hungry applications such as decision support, database management, data
mining, data warehousing, and e-commerce will see significant performance improvements
through the adoption of 64-bit solutions.
Applications requiring higher floating point capabilities, such as number crunching, analytics,
rendering, and high-performance technical computing (HPTC) will also see performance
advantages over 32-bit systems.
As compilers mature, many software vendors and HPTC
companies will be investigating IA-64 systems as an alternative to UNIX platforms.
Where the benefits of IA-64 can be recognized, many software vendors and companies with
custom code are using IA-64 systems as a development platform for porting their code.
Since any
code that is ported on IA-64 systems can easily be migrated to the follow-on platform based on the
next-generation 64-bit processor, it makes sense that these companies start early to increase their
competitiveness in the market.
3
TC020702TB