Datasheet

34
Product Guide
Mac OS X Tiger
64-Bit Computing
64-bit computing is the next big step in delivering greater computing power to solve
even the most challenging problems. It provides scientists, engineers, and other power
users with the tools to address problems that are billions of times larger than those
they can solve today with 32-bit computing.
Tiger extends the 64-bit capabilities of Mac OS X to allow applications to address vast
amounts of memory—up to four billion times the memory that can be addressed by
32-bit applications. In addition to 64-bit addressing, Tiger supports 64-bit math precision
and data throughput. And Tiger delivers the power and precision of 64-bit computing
while maintaining full performance and compatibility with existing 32-bit Mac OS X
applications.
Tiger makes it easy for developers to take advantage of 64-bit technology in Mac OS X
by providing a full set of 64-bit development tools as part of Xcode 2, as well as highly
optimized arithmetic and vector processing libraries. It’s even easy to port existing 64-bit
applications, because Tiger uses the same 64-bit data model as other UNIX-based systems.
Mac OS X Tiger offers the simplicity of Macintosh for personal productivity while deliv-
ering all of the 64-bit power required by the most demanding scientific, technical, and
creative applications.
New Possibilities
64-bit computing provides two important advances over 32-bit computing:
• 64-bit math operations. A wide range of applications, from bioinformatics to film and
video, make intensive use of complex math while requiring a high degree of precision.
The PowerPC G5 provides the ability to perform double-precision math operations in
a single clock cycle, as well as allowing more parallel computation for lower-precision
math operations.
• 64-bit memory addressing. 32-bit applications are limited to accessing 4 gigabytes
of data. 64-bit addressing makes up to four billion times as much memory available to
applications, providing access to as much memory as even the largest data sets require.
With a theoretical address space of 16 exabytes, Tiger is well poised to take advantage
of the next several generations in memory and storage.
How much memory do you get with 64-bit
computing? It’s important to understand that
64 bits isn’t just twice as good as 32 bits. The
amount of memory an application can use
doubles for each bit in the address, so a 33-bit
application could use twice as much memory
as a 32-bit application, a 34-bit application
twice as much as a 33-bit application, and
so on.
A 64-bit address space is over four billion
times larger than a 32-bit address space.
Imagine that a single address in a computer is
a spot roughly the size of a postage stamp. If
you crowded together as many of those spots
as you could fit into the memory of a 32-bit
application, they would cover an area roughly
equal to the roadway that spans the Golden
Gate Bridge. Now, if you crowded as many of
those spots together as you could fit into the
memory of a 64-bit application, they would
cover the entire surface of the earth.