Editing 3DNow! (section)

===3DNow!===
The first implementation of 3DNow! technology contains 21 new instructions that support [[SIMD]] floating-point operations. The 3DNow! data format is packed, [[single-precision]], floating-point. The 3DNow! instruction set also includes operations for SIMD integer operations, data prefetch, and faster MMX-to-floating-point switching. Later, [[Intel]] would add similar (but incompatible) instructions to the [[Pentium III]], known as [[Streaming SIMD Extensions|SSE]] (Streaming SIMD Extensions).

3DNow! floating-point instructions are the following:
{{Div col|colwidth=30em}}
* <code>PI2FD</code>{{snd}} Packed 32-bit integer to floating-point conversion
* <code>PF2ID</code>{{snd}} Packed floating-point to 32-bit integer conversion
* <code>PFCMPGE</code>{{snd}} Packed floating-point comparison, greater or equal
* <code>PFCMPGT</code>{{snd}} Packed floating-point comparison, greater
* <code>PFCMPEQ</code>{{snd}} Packed floating-point comparison, equal
* <code>PFACC</code>{{snd}} Packed floating-point accumulate
* <code>PFADD</code>{{snd}} Packed floating-point addition
* <code>PFSUB</code>{{snd}} Packed floating-point subtraction
* <code>PFSUBR</code>{{snd}} Packed floating-point reverse subtraction
* <code>PFMIN</code>{{snd}} Packed floating-point minimum
* <code>PFMAX</code>{{snd}} Packed floating-point maximum
* <code>PFMUL</code>{{snd}} Packed floating-point multiplication
* <code>PFRCP</code>{{snd}} Packed floating-point reciprocal approximation
* <code>PFRSQRT</code>{{snd}} Packed floating-point reciprocal square root approximation
* <code>PFRCPIT1</code>{{snd}} Packed floating-point reciprocal, first iteration step
* <code>PFRSQIT1</code>{{snd}} Packed floating-point reciprocal square root, first iteration step
* <code>PFRCPIT2</code>{{snd}} Packed floating-point reciprocal/reciprocal square root, second iteration step
{{div col end}}

3DNow! integer instructions are the following:
* <code>PAVGUSB</code>{{snd}} Packed 8-bit unsigned integer averaging
* <code>PMULHRW</code>{{snd}} Packed 16-bit integer multiply with rounding

3DNow! performance-enhancement instructions are the following:
* <code>FEMMS</code>{{snd}} Faster entry/exit of the MMX or floating-point state
* <code>PREFETCH/PREFETCHW</code>{{snd}} Prefetch at least a 32-byte line into L1 data cache (this is the only non-deprecated instruction)