<s>
Tachyon	B-Application
is	O
a	O
parallel/multiprocessor	O
ray	B-Algorithm
tracing	I-Algorithm
software	O
.	O
</s>
<s>
It	O
is	O
a	O
parallel	O
ray	B-Algorithm
tracing	I-Algorithm
library	O
for	O
use	O
on	O
distributed	O
memory	O
parallel	O
computers	O
,	O
shared	O
memory	O
computers	O
,	O
and	O
clusters	O
of	O
workstations	O
.	O
</s>
<s>
Tachyon	B-Application
implements	O
rendering	O
features	O
such	O
as	O
ambient	O
occlusion	O
lighting	O
,	O
depth-of-field	O
focal	O
blur	O
,	O
shadows	O
,	O
reflections	O
,	O
and	O
others	O
.	O
</s>
<s>
It	O
was	O
originally	O
developed	O
for	O
the	O
Intel	B-Device
iPSC/860	I-Device
by	O
John	O
Stone	O
for	O
his	O
M.S.	O
</s>
<s>
Tachyon	B-Application
subsequently	O
became	O
a	O
more	O
functional	O
and	O
complete	O
ray	B-Algorithm
tracing	I-Algorithm
engine	O
,	O
and	O
it	O
is	O
now	O
incorporated	O
into	O
a	O
number	O
of	O
other	O
open	O
source	O
software	O
packages	O
such	O
as	O
VMD	B-Application
,	O
and	O
SageMath	B-Application
.	O
</s>
<s>
Tachyon	B-Application
is	O
released	O
under	O
a	O
permissive	O
license	O
(	O
included	O
in	O
the	O
tarball	O
)	O
.	O
</s>
<s>
Tachyon	B-Application
was	O
originally	O
developed	O
for	O
the	O
Intel	B-Device
iPSC/860	I-Device
,	O
a	O
distributed	O
memory	O
parallel	O
computer	O
based	O
on	O
a	O
hypercube	O
interconnect	O
topology	O
based	O
on	O
the	O
Intel	O
i860	O
,	O
an	O
early	O
RISC	O
CPU	O
with	O
VLIW	O
architecture	O
and	O
.	O
</s>
<s>
Tachyon	B-Application
was	O
originally	O
written	O
using	O
Intel	O
's	O
proprietary	O
NX	O
message	O
passing	O
interface	O
for	O
the	O
iPSC	O
series	O
,	O
but	O
it	O
was	O
ported	O
to	O
the	O
earliest	O
versions	O
of	O
MPI	O
shortly	O
thereafter	O
in	O
1995	O
.	O
</s>
<s>
Tachyon	B-Application
was	O
adapted	O
to	O
run	O
on	O
the	O
Intel	B-Device
Paragon	I-Device
platform	O
using	O
the	O
at	O
Oak	O
Ridge	O
National	O
Laboratory	O
.	O
</s>
<s>
The	O
ORNL	O
XP/S	O
150	O
MP	O
was	O
the	O
first	O
platform	O
Tachyon	B-Application
supported	O
that	O
combined	O
both	O
large-scale	O
distributed	O
memory	O
message	O
passing	O
among	O
nodes	O
,	O
and	O
shared	O
memory	O
multithreading	O
within	O
nodes	O
.	O
</s>
<s>
Adaptation	O
of	O
Tachyon	B-Application
to	O
a	O
variety	O
of	O
conventional	O
Unix-based	O
workstation	O
platforms	O
and	O
early	O
clusters	O
followed	O
,	O
including	O
porting	O
to	O
the	O
IBM	O
SP2	O
.	O
</s>
<s>
Tachyon	B-Application
was	O
incorporated	O
into	O
the	O
PARAFLOW	O
CFD	O
code	O
to	O
allow	O
in-situ	O
volume	O
visualization	O
of	O
supersonic	O
combustor	O
flows	O
performed	O
on	O
the	O
Paragon	O
XP/S	O
at	O
NASA	O
Langley	O
Research	O
Center	O
,	O
providing	O
a	O
significant	O
performance	O
gain	O
over	O
conventional	O
post-processing	O
visualization	O
approaches	O
that	O
had	O
been	O
used	O
previously	O
.	O
</s>
<s>
Beginning	O
in	O
1999	O
,	O
support	O
for	O
Tachyon	B-Application
was	O
incorporated	O
into	O
the	O
molecular	O
graphics	O
program	O
VMD	B-Application
,	O
and	O
this	O
began	O
an	O
ongoing	O
period	O
co-development	O
of	O
Tachyon	B-Application
and	O
VMD	B-Application
where	O
many	O
new	O
Tachyon	B-Application
features	O
were	O
added	O
specifically	O
for	O
molecular	O
graphics	O
.	O
</s>
<s>
Tachyon	B-Application
was	O
used	O
to	O
render	O
the	O
winning	O
image	O
illustration	O
category	O
for	O
the	O
NSF	O
2004	O
Visualization	O
Challenge	O
.	O
</s>
<s>
In	O
2007	O
,	O
Tachyon	B-Application
added	O
support	O
for	O
ambient	O
occlusion	O
lighting	O
,	O
which	O
was	O
one	O
of	O
the	O
features	O
that	O
made	O
it	O
increasingly	O
popular	O
for	O
molecular	O
visualization	O
in	O
conjunction	O
with	O
VMD	B-Application
.	O
</s>
<s>
VMD	B-Application
and	O
Tachyon	B-Application
were	O
gradually	O
adapted	O
to	O
support	O
routine	O
visualization	O
and	O
analysis	O
tasks	O
on	O
clusters	O
,	O
and	O
later	O
for	O
large	O
petascale	O
supercomputers	O
.	O
</s>
<s>
Tachyon	B-Application
was	O
used	O
to	O
produce	O
figures	O
,	O
movies	O
,	O
and	O
the	O
Nature	O
of	O
the	O
atomic	O
structure	O
of	O
the	O
HIV-1	O
capsid	O
solved	O
by	O
Zhao	O
et	O
al	O
.	O
</s>
<s>
in	O
2013	O
,	O
on	O
the	O
Blue	B-Device
Waters	I-Device
petascale	I-Device
supercomputer	I-Device
at	O
NCSA	O
,	O
U	O
.	O
Illinois	O
.	O
</s>
<s>
Both	O
CPU	O
and	O
GPU	O
versions	O
of	O
Tachyon	B-Application
were	O
used	O
to	O
render	O
images	O
of	O
the	O
SARS-CoV-2	O
virion	O
,	O
spike	O
protein	O
,	O
and	O
aerosolized	O
virion	O
in	O
three	O
separate	O
ACM	O
Gordon	O
Bell	O
COVID-19	O
research	O
projects	O
,	O
including	O
the	O
winning	O
project	O
at	O
Supercomputing	O
2020	O
,	O
and	O
two	O
finalist	O
projects	O
at	O
Supercomputing	O
2021	O
.	O
</s>
<s>
Owing	O
in	O
part	O
to	O
its	O
portability	O
to	O
a	O
diverse	O
range	O
of	O
platforms	O
Tachyon	B-Application
has	O
been	O
used	O
as	O
a	O
test	O
case	O
for	O
a	O
variety	O
of	O
parallel	O
computing	O
and	O
compiler	O
research	O
articles	O
.	O
</s>
<s>
In	O
1999	O
,	O
John	O
Stone	O
assisted	O
Bill	O
Magro	O
with	O
adaptation	O
of	O
Tachyon	B-Application
to	O
support	O
early	O
versions	O
of	O
the	O
OpenMP	O
directive-based	O
parallel	O
computing	O
standard	O
,	O
using	O
Kuck	O
and	O
Associates	O
 '	O
KCC	O
compiler	O
.	O
</s>
<s>
Tachyon	B-Application
was	O
shown	O
as	O
a	O
demo	O
performing	O
interactive	O
ray	B-Algorithm
tracing	I-Algorithm
on	O
DEC	O
Alpha	O
workstations	O
using	O
KCC	O
and	O
OpenMP	O
.	O
</s>
<s>
In	O
2000	O
,	O
Intel	O
acquired	O
Kuck	O
and	O
Associates	O
Inc.	O
,	O
and	O
Tachyon	B-Application
continued	O
to	O
be	O
used	O
as	O
an	O
OpenMP	O
demonstration	O
.	O
</s>
<s>
Intel	O
later	O
used	O
Tachyon	B-Application
to	O
develop	O
a	O
variety	O
of	O
programming	O
examples	O
for	O
its	O
Threading	O
Building	O
Blocks	O
(	O
TBB	O
)	O
parallel	O
programming	O
system	O
,	O
where	O
an	O
old	O
version	O
of	O
the	O
program	O
continues	O
to	O
be	O
incorporated	O
as	O
an	O
example	O
to	O
the	O
present	O
day	O
.	O
</s>
<s>
In	O
2006	O
,	O
Tachyon	B-Application
was	O
selected	O
by	O
the	O
SPEC	O
HPG	O
for	O
inclusion	O
in	O
the	O
benchmark	O
suite	O
.	O
</s>
<s>
Beyond	O
Tachyon	B-Application
's	O
typical	O
use	O
as	O
tool	O
for	O
rendering	O
high	O
quality	O
images	O
,	O
likely	O
due	O
to	O
its	O
portability	O
and	O
inclusion	O
in	O
SPEC	O
MPI	O
2007	O
,	O
it	O
has	O
also	O
been	O
used	O
as	O
a	O
test	O
case	O
and	O
point	O
of	O
comparison	O
for	O
a	O
variety	O
of	O
research	O
projects	O
related	O
to	O
parallel	O
rendering	O
and	O
visualization	O
,	O
cloud	O
computing	O
,	O
and	O
parallel	O
computing	O
,	O
compilers	O
,	O
runtime	O
systems	O
,	O
and	O
computer	O
architecture	O
,	O
performance	O
analysis	O
tools	O
,	O
and	O
energy	O
efficiency	O
of	O
HPC	O
systems	O
.	O
</s>
