Cray CRAY-1 - 064 Ijk Floating Product of (Sj) and (Sk) to si; 065 Ijk Half-Precision Rounded Floating Product of (Sj) and; 066 Ijk Rounded Floating Product of (Sj) and (Sk) to si; 067 Ijk Reciprocal Iteration; 2 - (Sj) * (Sk) to si

r-------------------------------------------------------------------
064ijk
065ijk
Floating product of (Sj) and (Sk) to Si
Half-precision rounded floating product of (Sj) and
(Sk) to Si
066ijk
067ijk
Rounded floating product of (Sj) and (Sk) to Si
Reciprocal iteration; 2 - (Sj) * (Sk) to 5i
These instructions are executed by the floating point multiply unit.
Operands are assumed to be in floating point format. The result is
not guaranteed to be normalized if the operands are unnormalizec.
The 064 instruction forms the product of the floating point quantities
in Sj and Sk and enters the result into Si.
The 065 instruction forms the half-precision rounded product of the
floating point quantities in Sj and Sk and enters the result into Si.
The low order 18 bits of the result are cleared.
The 066 instruction forms the rounded product of the floating point
quantities in Sj and Sk and enters the result into Si.
The 067 instruction forms two minus the product of the floating point
quantities in Sj and Sk and enters the result into Si. This instruction
is used in the divide sequence as described in Section 3 under Floating
Point Arithmetic.
Hold issue conditions
034 ~ 037 in process
Exchange in process
S register access conflict
Si, Sj, or Sk reserved
160 - 167 in process; unit busy (VL) + 4 CPs
2240004
4-42
E