Difference between revisions of "Instruction Set/fmad"
From Mill Computing Wiki
m (Protected "Instruction Set/fmad": generated ([Edit=<protect-level-bot>] (indefinite) [Move=<protect-level-bot>] (indefinite))) | |||
| Line 1: | Line 1: | ||
{{DISPLAYTITLE:fmad}} | {{DISPLAYTITLE:fmad}} | ||
| − | <div style="font-size:80%;line-height:90%;margin-bottom:2em">[[Speculation| | + | <div style="font-size:80%;line-height:90%;margin-bottom:2em">[[Speculation|speculable]] [[Encoding|exu stream]] [[Decode|exu block]] [[Phasing|compute phase]] operation [[Domains|in the decimal floating point value domain]] [[Rounding|and rounds use current dynamic rounding mode]]<br /> |
'''native on:''' [[Cores/Decimal8|Decimal8]] [[Cores/Decimal16|Decimal16]] <br /> | '''native on:''' [[Cores/Decimal8|Decimal8]] [[Cores/Decimal16|Decimal16]] <br /> | ||
</div> | </div> | ||
Revision as of 09:31, 9 February 2015
speculable exu stream exu block compute phase operation in the decimal floating point value domain and rounds use current dynamic rounding mode
Decimal floating point fused multiply-add. As usual for those, it yields a higher precision than doing it separately, and is faster too. Uses current rounding behaviour.
operands: like Addd [dd:d]
Returns x*y+z on the belt.
encoding:
fmad(d x)
,
exuArgs(op arg0, op arg1)
| Core | In Slots | Latencies |
|---|---|---|
| Decimal8 | E0 E1 | d,d:d=7 dv,dv:dv=7 q,q:q=8 qv,qv:qv=8 |
| Decimal16 | E0 E1 | d,d:d=7 dv,dv:dv=7 q,q:q=8 qv,qv:qv=8 |
fmad(d x, d y, d z, d w) → d r0, d r1
operands: like Fmasd [dd:d]
This is a fused multiply-add-subtract. An excellent way to make full use of all Functional Units in the 2 Slots.
r0 is x*y+z*w
r1 is x*y-z*w
encoding:
fmad(d x, d y)
,
exuArgs(op arg0, op arg1)
| Core | In Slots | Latencies |
|---|---|---|
| Decimal8 | E0 | d,d:d,d=7,7 dv,dv:dv,dv=7,7 q,q:q,q=8,8 qv,qv:qv,qv=8,8 |
| Decimal16 | E0 | d,d:d,d=7,7 dv,dv:dv,dv=7,7 q,q:q,q=8,8 qv,qv:qv,qv=8,8 |
Instruction Set, alphabetical, Instruction Set by Category, Instruction Set, sortable, filterable
