308 Instruction Latencies Appendix C
25112 Rev. 3.06 September 2005
Software Optimization Guide for AMD64 Processors
FCOMIP ST, ST(i) DFh 11-110-xxx VectorPath FADD 3 3
FCOMP ST(i) D8h 11-011-xxx DirectPath FADD 2 1
FCOMP [mem32real] D8h mm-011-xxx DirectPath FADD 4
FCOMP [mem64real] DCh mm-011-xxx DirectPath FADD 4
FCOMPP DEh 11-011-001 DirectPath FADD 2
FCOS D9h 11-111-111 VectorPath - 92
FDECSTP D9h 11-110-110 DirectPath FADD/FMUL/
FSTORE
2
FDIV ST, ST(i) D8h 11-110-xxx DirectPath FMUL 16/20
/24
1, 6
FDIV ST(i), ST DCh 11-111-xxx DirectPath FMUL 16/20
/24
1, 6
FDIV [mem32real] D8h mm-110-xxx DirectPath FMUL 18/22
/26
6
FDIV [mem64real] DCh mm-110-xxx DirectPath FMUL 18/22
/26
6
FDIVP ST(i), ST DEh 11-111-xxx DirectPath FMUL 16/20
/24
1, 6
FDIVR ST, ST(i) D8h 11-110-xxx DirectPath FMUL 16/20
/24
1, 6
FDIVR ST(i), ST DCh 11-111-xxx DirectPath FMUL 16/20
/24
1, 6
FDIVR [mem32real] D8h mm-111-xxx DirectPath FMUL 18/22
/26
6
FDIVR [mem64real] DCh mm-111-xxx DirectPath FMUL 18/22
/26
6
FDIVRP DEh 11-110-001 DirectPath FMUL 16/20
/24
6
Table 15. x87 Floating-Point Instructions (Continued)
Syntax
Encoding
Decode
type
FPU
pipe(s)
Latency Note
First
byte
Second
byte
ModRM byte
Notes:
1. The last three bits of the ModRM byte select the stack entry ST(i).
2. These instructions have an effective latency as shown. However, these instructions generate an internal NOP
with a latency of two cycles but no related dependencies. These internal NOPs can be executed at a rate of
three per cycle and can use any of the three execution resources.
3. This is a VectorPath decoded operation that uses one execution pipe (one ROP).
4. There is additional latency associated with this instruction. “e” represents the difference between the exponents
of the divisor and the dividend. If “s” is the number of normalization shifts performed on the result, then
n = (s+1)/2 where (0 <= n <= 32).
5. The latency provided for this operation is the best-case latency.
6. The three latency numbers represent the latency values for precision control settings of single precision, double
precision, and extended precision, respectively.
