<s>
In	O
computer	B-General_Concept
science	I-General_Concept
,	O
the	O
precision	B-Architecture
of	O
a	O
numerical	O
quantity	O
is	O
a	O
measure	O
of	O
the	O
detail	O
in	O
which	O
the	O
quantity	O
is	O
expressed	O
.	O
</s>
<s>
It	O
is	O
related	O
to	O
precision	B-Architecture
in	I-Architecture
mathematics	I-Architecture
,	O
which	O
describes	O
the	O
number	O
of	O
digits	O
that	O
are	O
used	O
to	O
express	O
a	O
value	O
.	O
</s>
<s>
Of	O
these	O
,	O
octuple-precision	O
format	O
is	O
rarely	O
used	O
.	O
</s>
<s>
The	O
single	O
-	O
and	O
double-precision	O
formats	O
are	O
most	O
widely	O
used	O
and	O
supported	O
on	O
nearly	O
all	O
platforms	O
.	O
</s>
<s>
The	O
use	O
of	O
half-precision	O
format	O
has	O
been	O
increasing	O
especially	O
in	O
the	O
field	O
of	O
machine	O
learning	O
since	O
many	O
machine	O
learning	O
algorithms	O
are	O
inherently	O
error-tolerant	O
.	O
</s>
<s>
Precision	B-Architecture
is	O
often	O
the	O
source	O
of	O
rounding	B-Algorithm
errors	I-Algorithm
in	O
computation	O
.	O
</s>
<s>
An	O
example	O
would	O
be	O
to	O
store	O
"	O
sin( 	O
0.1	O
)	O
"	O
in	O
IEEE	O
single	O
precision	B-Architecture
floating	O
point	O
standard	O
.	O
</s>
