<s>
The	O
Cheyenne	B-Device
supercomputer	O
at	O
the	O
NCAR-Wyoming	O
Supercomputing	O
Center	O
(	O
NWSC	O
)	O
in	O
Cheyenne	B-Device
,	O
Wyoming	O
began	O
operation	O
as	O
one	O
of	O
the	O
world	O
’s	O
most	O
powerful	O
and	O
energy-efficient	O
computers	O
.	O
</s>
<s>
Ranked	O
in	O
November	O
2016	O
as	O
the	O
20th	O
most	O
powerful	O
computer	O
in	O
the	O
world	O
by	O
Top500	B-Operating_System
,	O
the	O
5.34-petaflops	O
system	O
is	O
capable	O
of	O
more	O
than	O
triple	O
the	O
amount	O
of	O
scientific	O
computing	O
performed	O
by	O
NCAR	O
’s	O
previous	O
supercomputer	O
,	O
Yellowstone	B-Device
.	O
</s>
<s>
It	O
also	O
is	O
three	O
times	O
more	O
energy	O
efficient	O
than	O
Yellowstone	B-Device
,	O
with	O
a	O
peak	O
computation	O
rate	O
of	O
more	O
than	O
3	O
billion	O
calculations	O
per	O
second	O
for	O
every	O
watt	O
of	O
energy	O
consumed	O
.	O
</s>
<s>
The	O
National	O
Science	O
Foundation	O
and	O
the	O
State	O
of	O
Wyoming	O
through	O
an	O
appropriation	O
to	O
the	O
University	O
of	O
Wyoming	O
funded	O
Cheyenne	B-Device
to	O
provide	O
the	O
United	O
States	O
with	O
a	O
major	O
new	O
tool	O
to	O
advance	O
understanding	O
of	O
the	O
atmospheric	O
and	O
related	O
Earth	O
system	O
sciences	O
.	O
</s>
<s>
High-performance	O
computers	O
such	O
as	O
Cheyenne	B-Device
allow	O
researchers	O
to	O
run	O
increasingly	O
detailed	O
models	O
that	O
simulate	O
complex	O
processes	O
to	O
estimate	O
how	O
they	O
might	O
unfold	O
in	O
the	O
future	O
.	O
</s>
<s>
Cheyenne	B-Device
’s	O
users	O
advance	O
the	O
knowledge	O
needed	O
for	O
saving	O
lives	O
,	O
protecting	O
property	O
,	O
and	O
enabling	O
U.S.	O
businesses	O
to	O
better	O
compete	O
in	O
the	O
global	O
marketplace	O
.	O
</s>
<s>
Scientists	O
across	O
the	O
country	O
will	O
use	O
Cheyenne	B-Device
to	O
study	O
phenomena	O
ranging	O
from	O
weather	O
and	O
climate	O
to	O
wildfires	O
,	O
seismic	O
activity	O
,	O
and	O
airflows	O
that	O
generate	O
power	O
at	O
wind	O
farms	O
.	O
</s>
<s>
The	O
supercomputer	O
’s	O
name	O
was	O
chosen	O
to	O
honor	O
the	O
people	O
of	O
Cheyenne	B-Device
,	O
Wyoming	O
,	O
who	O
supported	O
the	O
installation	O
of	O
the	O
NWSC	O
and	O
its	O
computers	O
there	O
.	O
</s>
<s>
The	O
name	O
also	O
commemorates	O
the	O
150th	O
anniversary	O
of	O
the	O
city	O
,	O
which	O
was	O
founded	O
in	O
1867	O
and	O
named	O
for	O
the	O
Native	O
American	O
Cheyenne	B-Device
Nation	O
.	O
</s>
<s>
The	O
Cheyenne	B-Device
supercomputer	O
was	O
built	O
by	O
Silicon	O
Graphics	O
International	O
Corporation	O
(	O
SGI	O
)	O
in	O
coordination	O
with	O
centralized	O
file	B-Application
system	I-Application
and	O
data	O
storage	O
components	O
provided	O
by	O
DataDirect	O
Networks	B-Architecture
(	O
DDN	O
)	O
.	O
</s>
<s>
The	O
new	O
data	O
storage	O
system	O
for	O
Cheyenne	B-Device
is	O
integrated	O
with	O
NCAR	O
’s	O
existing	O
GLADE	O
file	B-Application
system	I-Application
.	O
</s>
<s>
Cheyenne	B-Device
is	O
an	O
SGI	O
ICE	O
XA	O
system	O
with	O
4,032	O
dual-socket	O
scientific	O
computation	O
nodes	O
running	O
18-core	O
2.3-GHz	O
Intel	O
Xeon	O
E5-2697v4	O
processors	O
with	O
203	O
[	O
now	O
315 ]	O
terabytes	O
of	O
memory	O
.	O
</s>
<s>
Cheyenne	B-Device
runs	O
the	O
SUSE	O
Linux	O
Enterprise	O
Server	O
12	O
SP1	O
operating	O
system	O
.	O
</s>
<s>
Cheyenne	B-Device
is	O
integrated	O
with	O
many	O
other	O
high-performance	O
computing	O
resources	O
in	O
the	O
NWSC	O
.	O
</s>
<s>
The	O
central	O
feature	O
of	O
this	O
supercomputing	O
architecture	O
is	O
its	O
shared	O
file	B-Application
system	I-Application
that	O
streamlines	O
science	O
workflows	B-Operating_System
by	O
providing	O
computation	O
,	O
analysis	O
,	O
and	O
visualization	B-Application
work	O
spaces	O
common	O
to	O
all	O
resources	O
.	O
</s>
<s>
This	O
common	O
data	O
storage	O
pool	O
,	O
called	O
the	O
GLobally	O
Accessible	O
Data	O
Environment	O
(	O
GLADE	O
)	O
,	O
provides	O
36.4	O
petabytes	O
of	O
online	O
disk	O
capacity	O
shared	O
by	O
the	O
supercomputers	O
,	O
two	O
data	B-General_Concept
analysis	I-General_Concept
and	O
visualization	B-Application
(	O
DAV	O
)	O
cluster	O
computers	O
,	O
data	O
servers	O
for	O
both	O
local	O
and	O
remote	O
users	O
,	O
and	O
a	O
data	O
archive	O
with	O
the	O
capacity	O
to	O
store	O
320	O
petabytes	O
of	O
research	O
data	O
.	O
</s>
<s>
High-speed	O
networks	B-Architecture
connect	O
this	O
Cheyenne	B-Device
environment	O
to	O
science	O
gateways	O
,	O
data	O
transfer	O
services	O
,	O
remote	O
visualization	B-Application
resources	O
,	O
Extreme	O
Science	O
and	O
Engineering	O
Discovery	O
Environment	O
(	O
XSEDE	O
)	O
sites	O
,	O
and	O
partner	O
sites	O
around	O
the	O
world	O
.	O
</s>
<s>
This	O
integration	O
of	O
computing	O
resources	O
,	O
file	B-Application
systems	I-Application
,	O
data	O
storage	O
,	O
and	O
broadband	O
networks	B-Architecture
allows	O
scientists	O
to	O
simulate	O
future	O
geophysical	O
scenarios	O
at	O
high	O
resolution	O
,	O
then	O
analyze	O
and	O
visualize	O
them	O
on	O
one	O
computing	O
complex	O
.	O
</s>
<s>
Cheyenne	B-Device
makes	O
more	O
than	O
1.2	O
billion	O
core-hours	O
available	O
each	O
year	O
to	O
researchers	O
in	O
the	O
Earth	O
system	O
sciences	O
.	O
</s>
