Intel i86W Manual page 174

PROGRAMMING EXAMPLES
II
PIPELINED DUAL-OPERATION INSTRUCTION DOUBLE PRECISION
II
Multiplier Adder
II
Stages Stages
II
1 2 1 2
3 Result
II Priming phase
m12apm.dd f4, f16,f0
II
1*6 ?? ?? ?? ?? Discard
m12apm.dd f6, f18,f0 II
2*5
1*6 ?? ??
?? Discard
pfadd.dd f0, f0 ,f0 II 0 ?? ?? Discard
pfadd.dd f0, f0 ,f0 II
0 0
?? Discard
pfadd.dd f0, f0 ,f0
II
0
0
0
Discard
II Continuous operation phase
m12apm.dd f8, f20,f0 II 3*4
2*5
6+0
0 0 Discard
m12apm.dd f10,f22,f0 II 4*3
3*4
10+0 6+0
0
Discard
m12apm.dd f12,f24,f0
II 5*2
4*3
12+0 10+0
6 Discard
m12apm.dd f14,f26,f0
II 6*1
5*2
12+6 12+0 10 Discard
II
For larger vectors, include more instructions here
II
Flushing phase
m12apm.dd f0, f0, f0 II 0*0
6*1
10+10 12+6 12 Discard
m12apm.dd f0, f0, f0
II 0*0
0*0
6+12 10+10
18 Discard
II
Three partial sums are now in the adder pipeline.
pfadd.dd f0 ,f0 ,f28 II 0 6+12 20 f28
=
18
pfadd.dd f28,f30,f30
II 18+20
0 18 f30
=
20
pfadd.dd f0 ,f0 ,f28 II
0
18+20 0 f28
=
18
pfadd.dd f0 ,f0 ,f0
II
0 0 38 Discard
pfadd.dd f0 ,f0 ,f30 II
0 0 0
f30 38
fadd.dd f28,f30,f30
II
f30
=
56
Example 9-11. Pipelined Double-Precision Dual Operation
9-12