AMD amd64 architecture User Manual
64-Bit Media Programming
199
24592—Rev. 3.15—November 2009
AMD64 Technology
5.3.6 Floating-Point (3DNow!™) Vector Operations
Floating-point vector instructions using the MMX registers were introduced by AMD with the
3DNow! technology. These instructions take 64-bit vector operands consisting of two 32-bit single-
precision floating-point numbers, shown as FP single in Figure 5-6.
3DNow! technology. These instructions take 64-bit vector operands consisting of two 32-bit single-
precision floating-point numbers, shown as FP single in Figure 5-6.
Figure 5-6.
Floating-Point (3DNow!™ Instruction) Operations
The AMD64 architecture’s 3DNow! floating-point instructions provide a unique advantage over
legacy x87 floating-point instructions: They allow integer and floating-point instructions to be
intermixed in the same procedure, using only the MMX registers. This avoids the need to switch
between integer MMX procedures and x87 floating-point procedures—a switch that may involve
time-consuming state saves and restores—while at the same time leaving the 128-bit XMM register
resources free for other applications.
legacy x87 floating-point instructions: They allow integer and floating-point instructions to be
intermixed in the same procedure, using only the MMX registers. This avoids the need to switch
between integer MMX procedures and x87 floating-point procedures—a switch that may involve
time-consuming state saves and restores—while at the same time leaving the 128-bit XMM register
resources free for other applications.
The 3DNow! instructions allow applications such as 3D graphics to accelerate front-end geometry,
clipping, and lighting calculations. Picture and pixel data are typically integer data types, although
both integer and floating-point instructions are often required to operate completely on the data. For
example, software can change the viewing perspective of a 3D scene through transformation matrices
by using floating-point instructions in the same procedure that contains integer operations on other
aspects of the graphics data.
clipping, and lighting calculations. Picture and pixel data are typically integer data types, although
both integer and floating-point instructions are often required to operate completely on the data. For
example, software can change the viewing perspective of a 3D scene through transformation matrices
by using floating-point instructions in the same procedure that contains integer operations on other
aspects of the graphics data.
3DNow! programs typically perform better than x87 floating-point code, because the MMX register
file is flat rather than stack-oriented and because 3DNow! instructions can operate on twice as many
operands as x87 floating-point instructions. This ability to operate in parallel on twice as many
floating-point values in the same register space often makes it possible to remove local temporary
variables in 3DNow! code that would otherwise be needed in x87 floating-point code.
file is flat rather than stack-oriented and because 3DNow! instructions can operate on twice as many
operands as x87 floating-point instructions. This ability to operate in parallel on twice as many
floating-point values in the same register space often makes it possible to remove local temporary
variables in 3DNow! code that would otherwise be needed in x87 floating-point code.
63
31
0
32
FP single FP single
63
31
0
32
FP single FP single
63
31
0
32
FP single FP single
513-124.eps
63
0
63
0
op
op