4.6.4
Approximation
4.6.4.1
Floating-point Reciprocal Approximation
There are two Reciprocal Approximation instructions. The first, in major op 0, encodes
the full register variant. The second, in major op 1, encodes the parallel variant.
40
F6
Instruction
frcpa.sf
fprcpa.sf
4.6.4.2
Floating-point Reciprocal Square Root Approximation
There are two Reciprocal Square Root Approximation instructions. The first, in major op
0, encodes the full register variant. The second, in major op 1, encodes the parallel
variant.
40
F7
Instruction
frsqrta.sf
fprsqrta.sf
Volume 3: Instruction Formats
37 36 35 34 33 32
0 - 1
q sf x
p
2
4
1
2
1
6
Operands
f
, p
= f
, f
1
2
2
3
37 36 35 34 33 32
0 - 1
q sf x
p
2
4
1
2
1
6
Operands
f
, p
= f
1
2
3
27 26
20 19
f
f
3
2
7
7
Opcode
x
0
1
1
27 26
20 19
f
3
7
7
Opcode
x
0
1
1
13 12
6 5
f
1
7
Extension
q
sf
See Table 4-63 on
0
page 3:358
13 12
6 5
f
1
7
Extension
q
sf
See Table 4-63 on
1
page 3:358
0
qp
6
0
qp
6
3:361