1
00:00:00,040 --> 00:00:02,480
The following content is
provided under a Creative

2
00:00:02,480 --> 00:00:04,000
Commons license.

3
00:00:04,000 --> 00:00:06,340
Your support will help
MIT OpenCourseWare

4
00:00:06,340 --> 00:00:10,710
continue to offer high quality
educational resources for free.

5
00:00:10,710 --> 00:00:13,320
To make a donation or
view additional materials

6
00:00:13,320 --> 00:00:17,190
from hundreds of MIT courses,
visit MIT OpenCourseWare

7
00:00:17,190 --> 00:00:17,815
at ocw.mit.edu.

8
00:00:22,000 --> 00:00:24,000
PRAMOOK KHUNGUM: Hi, my
name is Pramook Khungum.

9
00:00:24,000 --> 00:00:26,100
I'm supposed to be one
of four, but it remains

10
00:00:26,100 --> 00:00:29,200
that I am the only one now.

11
00:00:29,200 --> 00:00:33,100
So my project was about
global illumination.

12
00:00:33,100 --> 00:00:37,230
And it's about
rendering 3D scenes.

13
00:00:37,230 --> 00:00:40,700
So first, let me tell you about
the direct illumination model.

14
00:00:40,700 --> 00:00:45,020
This is something you can
achieve with ray tracing.

15
00:00:45,020 --> 00:00:47,760
As you can see,
the image above is

16
00:00:47,760 --> 00:00:51,310
rendered by shooting
a ray from your eyes,

17
00:00:51,310 --> 00:00:54,780
and then taking
the local lighting

18
00:00:54,780 --> 00:00:57,270
properties of the
materials, and then

19
00:00:57,270 --> 00:01:00,020
see whether there's a light
coming to that point or not,

20
00:01:00,020 --> 00:01:01,990
and then figure out the color.

21
00:01:01,990 --> 00:01:05,290
So in this model, there's a lot
things unnatural about this.

22
00:01:05,290 --> 00:01:08,090
Because as we can see, the
light is an area light,

23
00:01:08,090 --> 00:01:11,650
but then the shadow
is very solid shadow

24
00:01:11,650 --> 00:01:14,960
as if it's lit by a point
light or something like that.

25
00:01:14,960 --> 00:01:19,260
And you can see that the
ceiling is not lit at all.

26
00:01:19,260 --> 00:01:21,980
This means that this
model of lighting

27
00:01:21,980 --> 00:01:23,850
doesn't take into
account the light that

28
00:01:23,850 --> 00:01:26,630
bounces between the
walls and lighting

29
00:01:26,630 --> 00:01:35,830
parts that should have been
lit in the real life scene,

30
00:01:35,830 --> 00:01:36,740
as you can see.

31
00:01:36,740 --> 00:01:41,460
So global illumination
basically tries to fix that.

32
00:01:41,460 --> 00:01:45,340
And the effects you can achieve
with global illumination

33
00:01:45,340 --> 00:01:50,850
algorithms are
basically soft shadows,

34
00:01:50,850 --> 00:01:53,240
color bleeding-- you
can see that there's

35
00:01:53,240 --> 00:01:57,110
a very pale shades
of blue here--

36
00:01:57,110 --> 00:02:00,200
and caustics, which means
that light from here

37
00:02:00,200 --> 00:02:05,180
goes through the glasses
and then focus on here.

38
00:02:05,180 --> 00:02:09,750
So at first when I
started the project,

39
00:02:09,750 --> 00:02:12,440
I wanted to globally illuminate
a very simple scene, which

40
00:02:12,440 --> 00:02:14,090
is the Cornell Box.

41
00:02:14,090 --> 00:02:18,920
So it's basically this box, and
you replace these two spheres

42
00:02:18,920 --> 00:02:20,634
with two white boxes.

43
00:02:20,634 --> 00:02:22,800
And then you can observe
color bleeding in something

44
00:02:22,800 --> 00:02:25,360
like that.

45
00:02:25,360 --> 00:02:30,090
And I wanted to do it in
real time with caustics.

46
00:02:30,090 --> 00:02:35,980
However, the things I
accomplished are-- well,

47
00:02:35,980 --> 00:02:39,620
I globally illuminated the
Cornell Box as I promised.

48
00:02:39,620 --> 00:02:42,560
In real time--
no, it's too slow.

49
00:02:42,560 --> 00:02:46,580
With caustic, didn't
implement photon mapping.

50
00:02:46,580 --> 00:02:49,380
And sadly, there's still a
lot of room for optimization.

51
00:02:49,380 --> 00:02:54,740
So kind of disappointed
myself a little bit, but yeah,

52
00:02:54,740 --> 00:02:56,910
that's fine.

53
00:02:56,910 --> 00:03:03,080
So the algorithm I used
for global illumination

54
00:03:03,080 --> 00:03:04,820
is called Instant Radiosity.

55
00:03:04,820 --> 00:03:08,500
It's by some German guy
called Alexander Keller.

56
00:03:08,500 --> 00:03:11,260
It's good for scenes with
diffuse objects, which

57
00:03:11,260 --> 00:03:13,860
means that scenes
that doesn't have

58
00:03:13,860 --> 00:03:18,630
objects that reflect light or
lets light transmit through.

59
00:03:18,630 --> 00:03:23,360
Diffuse objects can be just this
table and something like that.

60
00:03:23,360 --> 00:03:25,620
Basically, it means that
when light strikes it,

61
00:03:25,620 --> 00:03:29,220
it just bounces off randomly.

62
00:03:29,220 --> 00:03:32,520
As such, it can't do
caustics, although it

63
00:03:32,520 --> 00:03:35,900
is very simple to implement
once you have the core.

64
00:03:35,900 --> 00:03:37,740
So let me describe
the algorithm.

65
00:03:37,740 --> 00:03:39,640
You start with the
light source, and then

66
00:03:39,640 --> 00:03:43,260
you emit artificial
light particles

67
00:03:43,260 --> 00:03:46,860
and let them bounce
around the scene.

68
00:03:46,860 --> 00:03:49,650
And then you will have
some more light source.

69
00:03:49,650 --> 00:03:52,490
And then you ray trace
using those light sources

70
00:03:52,490 --> 00:03:55,830
in the scene basically.

71
00:03:55,830 --> 00:03:58,600
So this is the illustration
of the algorithm.

72
00:03:58,600 --> 00:04:01,730
First, you have the
light, which goes off

73
00:04:01,730 --> 00:04:05,570
from this square light.

74
00:04:05,570 --> 00:04:08,330
And then you illuminate
the scene with it.

75
00:04:08,330 --> 00:04:09,830
And then you bounce
some lights off,

76
00:04:09,830 --> 00:04:11,500
and then illuminate the scene.

77
00:04:11,500 --> 00:04:13,610
Then you combine the
results, basically.

78
00:04:16,380 --> 00:04:21,200
So in my implementation,
at the beginning,

79
00:04:21,200 --> 00:04:24,410
I have one SPU scatter
all the light particles.

80
00:04:24,410 --> 00:04:29,080
The number of light particles
I scattered is about 128.

81
00:04:29,080 --> 00:04:31,510
And as they bounce
around, it ends up

82
00:04:31,510 --> 00:04:35,850
to be about 250 light
sources in total.

83
00:04:35,850 --> 00:04:40,510
And then I rendered a scene
by splitting the scene

84
00:04:40,510 --> 00:04:46,780
into horizontal slabs, and then
just give the slabs to each SPU

85
00:04:46,780 --> 00:04:48,050
to render.

86
00:04:48,050 --> 00:04:51,340
In this way, since the
scene is very small,

87
00:04:51,340 --> 00:04:55,080
it can just preload
them to every SPU.

88
00:04:55,080 --> 00:04:57,060
And they all can
work independently.

89
00:04:57,060 --> 00:05:02,360
So there's minimal
communication going around.

90
00:05:02,360 --> 00:05:08,479
And since I didn't implement
any other smart data structure

91
00:05:08,479 --> 00:05:10,020
to take care of the
geometry, it just

92
00:05:10,020 --> 00:05:11,620
used bruteforce ray tracing.

93
00:05:11,620 --> 00:05:14,752
My scenes have only
eight objects anyway.

94
00:05:14,752 --> 00:05:16,766
I wasn't maybe it
might be a good point

95
00:05:16,766 --> 00:05:18,224
to see if there's
[? a question. ?]

96
00:05:21,210 --> 00:05:24,376
AUDIENCE: So the SPUs are
doing rendering or ray tracing?

97
00:05:24,376 --> 00:05:26,250
PRAMOOK KHUNGUM: Ray
tracing, yes, basically.

98
00:05:26,250 --> 00:05:29,480
AUDIENCE: But I mean, can a ray
move from one slab to another

99
00:05:29,480 --> 00:05:32,390
depending on reflection
or something?

100
00:05:32,390 --> 00:05:34,030
PRAMOOK KHUNGUM: No.

101
00:05:34,030 --> 00:05:36,690
Basically, if you
trace the light there,

102
00:05:36,690 --> 00:05:38,730
the SPU has all the
ascription of the scene,

103
00:05:38,730 --> 00:05:41,110
so it can trace the ray
throughout the scene.

104
00:05:41,110 --> 00:05:44,405
But it's responsible for
only the pixels in the slab.

105
00:05:44,405 --> 00:05:45,214
AUDIENCE: Oh, OK.

106
00:05:47,940 --> 00:05:50,280
PRAMOOK KHUNGUM: The
complexity is kind of bad

107
00:05:50,280 --> 00:05:56,580
because the time is
you have your image

108
00:05:56,580 --> 00:06:00,200
size, the number of pixels times
the number of light particles,

109
00:06:00,200 --> 00:06:01,570
which is about 200.

110
00:06:01,570 --> 00:06:04,720
And then since I used
bruteforce ray shooting,

111
00:06:04,720 --> 00:06:07,750
you have to times by
the size of objects.

112
00:06:07,750 --> 00:06:10,710
But at first, I thought
I have only eight objects

113
00:06:10,710 --> 00:06:15,350
anyway so it's not going to
be a big performance rack.

114
00:06:15,350 --> 00:06:17,200
But then I came
to the realization

115
00:06:17,200 --> 00:06:20,660
that this algorithm
initially was conceived

116
00:06:20,660 --> 00:06:25,740
for using with programmable
graphics hardware, which

117
00:06:25,740 --> 00:06:29,260
can do really dumb tasks like
taking a point light source

118
00:06:29,260 --> 00:06:32,910
and illuminating a scene
with shadows very fast.

119
00:06:32,910 --> 00:06:36,250
So it turns out that this
algorithm is not really suited

120
00:06:36,250 --> 00:06:37,730
for the [? Cell, ?] basically.

121
00:06:40,750 --> 00:06:43,476
I wish I could have
done better on this.

122
00:06:43,476 --> 00:06:51,180
This 512 times 512 scene
took 20 seconds to render

123
00:06:51,180 --> 00:06:55,300
on the [? Cell. ?] But
the original paper, they

124
00:06:55,300 --> 00:06:57,270
rendered something
a little bigger

125
00:06:57,270 --> 00:07:00,981
than this in 24 seconds
with GPU basically,

126
00:07:00,981 --> 00:07:02,230
so it's kind of disappointing.

127
00:07:02,230 --> 00:07:03,750
PROFESSOR SAMAN: [INAUDIBLE].

128
00:07:03,750 --> 00:07:05,460
PRAMOOK KHUNGUM: Hello?

129
00:07:05,460 --> 00:07:08,866
PROFESSOR SAMAN: 24 seconds
on what type of a machine?

130
00:07:08,866 --> 00:07:10,640
Can you hear me?

131
00:07:10,640 --> 00:07:12,830
PRAMOOK KHUNGUM: Yes,
but not very clearly.

132
00:07:12,830 --> 00:07:15,199
AUDIENCE: 24 seconds on
what kind of a machine?

133
00:07:15,199 --> 00:07:18,060
PROFESSOR SAMAN: The original
paper took 24 seconds?

134
00:07:18,060 --> 00:07:23,200
PRAMOOK KHUNGUM: 24 seconds on
a 75 megahertz something machine

135
00:07:23,200 --> 00:07:24,779
with a GPU.

136
00:07:24,779 --> 00:07:25,612
PROFESSOR SAMAN: OK.

137
00:07:25,612 --> 00:07:30,910
And did you try without using
any SPU, just using CPU?

138
00:07:30,910 --> 00:07:35,130
PRAMOOK KHUNGUM: Just using the
CPU and the GPU and that one.

139
00:07:35,130 --> 00:07:37,130
PROFESSOR: His question
was, did you try using--

140
00:07:37,130 --> 00:07:40,106
PROFESSOR SAMAN: If you just
use CPU without any parallelism,

141
00:07:40,106 --> 00:07:42,590
how much time does it take?

142
00:07:42,590 --> 00:07:44,285
PRAMOOK KHUNGUM:
Oh, times that by 6.

143
00:07:48,400 --> 00:07:52,650
However, the speedup is
proportional to the number

144
00:07:52,650 --> 00:07:54,580
of processors you used.

145
00:07:54,580 --> 00:07:58,492
I basically graphed the runtime
with the number of processors.

146
00:07:58,492 --> 00:07:59,700
This is in [INAUDIBLE] scale.

147
00:07:59,700 --> 00:08:03,270
So as you can see,
it-- so that's

148
00:08:03,270 --> 00:08:08,588
the best you can do about
parallelism there, basically.

149
00:08:08,588 --> 00:08:10,420
AUDIENCE: [INAUDIBLE].

150
00:08:10,420 --> 00:08:12,170
PRAMOOK KHUNGUM: Yes?

151
00:08:12,170 --> 00:08:16,594
AUDIENCE: So runtime is
normalized to just the PPU?

152
00:08:16,594 --> 00:08:17,885
PRAMOOK KHUNGUM: Using the SPU.

153
00:08:20,978 --> 00:08:23,620
AUDIENCE: So what are
the units for the y-axis?

154
00:08:23,620 --> 00:08:27,169
PRAMOOK KHUNGUM: The y-axis
is runtime in seconds.

155
00:08:27,169 --> 00:08:27,835
AUDIENCE: I see.

156
00:08:27,835 --> 00:08:31,840
So 3.7 seconds, 3.6
seconds for one SPU?

157
00:08:31,840 --> 00:08:32,760
PRAMOOK KHUNGUM: Yes.

158
00:08:32,760 --> 00:08:36,070
3.6 seconds per one SPU.

159
00:08:36,070 --> 00:08:39,990
And as you increase,
the times go down.

160
00:08:39,990 --> 00:08:42,980
You just divide that by that.

161
00:08:42,980 --> 00:08:45,265
Is that clear?

162
00:08:45,265 --> 00:08:48,280
AUDIENCE: And they're all
using the PPU as well?

163
00:08:48,280 --> 00:08:50,485
PRAMOOK KHUNGUM: No, the
PPU doesn't do anything.

164
00:08:50,485 --> 00:08:51,085
I just used--

165
00:08:51,085 --> 00:08:53,710
AUDIENCE: Sorry, did you say 20
seconds on your previous slide?

166
00:08:53,710 --> 00:08:55,959
PRAMOOK KHUNGUM: This is
different [? image ?] slides.

167
00:08:55,959 --> 00:08:56,834
AUDIENCE: 128 by 128.

168
00:08:56,834 --> 00:08:57,708
PRAMOOK KHUNGUM: Yes.

169
00:08:57,708 --> 00:08:58,525
AUDIENCE: Oh, OK.

170
00:08:58,525 --> 00:08:59,170
OK, different--

171
00:08:59,170 --> 00:09:01,211
PRAMOOK KHUNGUM: Sorry,
I didn't make that clear.

172
00:09:01,211 --> 00:09:02,392
AUDIENCE: Sure, OK.

173
00:09:02,392 --> 00:09:03,600
PRAMOOK KHUNGUM: This is 128.

174
00:09:03,600 --> 00:09:07,540
So it's about 16 times faster.

175
00:09:07,540 --> 00:09:08,945
I couldn't wait for llike--

176
00:09:08,945 --> 00:09:09,290
AUDIENCE: Sure.

177
00:09:09,290 --> 00:09:10,206
PRAMOOK KHUNGUM: Yeah.

178
00:09:13,340 --> 00:09:16,690
So here are some
images I rendered.

179
00:09:16,690 --> 00:09:19,622
There are eight objects in here.

180
00:09:19,622 --> 00:09:23,690
And it took like 20.5
seconds with six SPUs.

181
00:09:26,200 --> 00:09:32,430
Another image that I rendered
is a miniature of a Cornell Box.

182
00:09:32,430 --> 00:09:38,225
It took 20.2 seconds
and six SPUs.

183
00:09:38,225 --> 00:09:40,920
AUDIENCE: Why is the
ceiling still so in black?

184
00:09:40,920 --> 00:09:42,928
PRAMOOK KHUNGUM: Yes, I
don't know that as well.

185
00:09:46,540 --> 00:09:51,290
Maybe it's about
the distribution

186
00:09:51,290 --> 00:09:55,320
of the particle light source.

187
00:09:55,320 --> 00:10:01,980
Probably I think when it goes
to the wall or to the floor,

188
00:10:01,980 --> 00:10:08,290
it gets severely attenuated
that it cannot really light

189
00:10:08,290 --> 00:10:10,810
the ceiling basically.

190
00:10:10,810 --> 00:10:16,100
So there's a comparison
between my rendering

191
00:10:16,100 --> 00:10:18,540
and the real Cornell Box.

192
00:10:18,540 --> 00:10:22,250
Basically, yes, there are still
a lot of [? boxes ?] around.

193
00:10:22,250 --> 00:10:24,540
Basically, the
ceiling is not lit.

194
00:10:24,540 --> 00:10:30,250
And it's kind of strange that
I see some gray tinges here

195
00:10:30,250 --> 00:10:33,460
instead of green as in this one.

196
00:10:33,460 --> 00:10:35,570
So probably, I have
to fix that as well.

197
00:10:35,570 --> 00:10:39,460
But as you can see, we have
achieved soft shadows and then

198
00:10:39,460 --> 00:10:40,974
some color bleeding.

199
00:10:40,974 --> 00:10:42,970
AUDIENCE: How do you
determine the location

200
00:10:42,970 --> 00:10:45,440
of the light sources?

201
00:10:45,440 --> 00:10:47,170
PRAMOOK KHUNGUM: How
do I determine the--

202
00:10:47,170 --> 00:10:47,961
AUDIENCE: Are you--

203
00:10:50,114 --> 00:10:51,613
PROFESSOR: Are you
giving this input

204
00:10:51,613 --> 00:10:53,790
to the algorithm, the
location of the light source?

205
00:10:53,790 --> 00:10:54,700
PRAMOOK KHUNGUM: Yes, basically.

206
00:10:54,700 --> 00:10:56,420
AUDIENCE: And they
can be arbitrary

207
00:10:56,420 --> 00:10:58,980
positions within the box or?

208
00:10:58,980 --> 00:11:01,690
PRAMOOK KHUNGUM: Yes, basically.

209
00:11:01,690 --> 00:11:04,830
This one is just
a square object,

210
00:11:04,830 --> 00:11:07,640
which is written in the
scene description file.

211
00:11:07,640 --> 00:11:12,760
And I can locate them
wherever I want basically.

212
00:11:12,760 --> 00:11:15,610
And the light particles
goes down from here.

213
00:11:23,800 --> 00:11:27,250
So there are some things that
can be further optimized.

214
00:11:27,250 --> 00:11:30,320
Basically, if I had decided
to use accelerated data

215
00:11:30,320 --> 00:11:36,980
structure, the complexity of
that in that size times objects

216
00:11:36,980 --> 00:11:41,290
times the number
of light sources,

217
00:11:41,290 --> 00:11:43,390
this term will be
reduced to [? log. ?]

218
00:11:43,390 --> 00:11:47,520
And probably, you
see about 3x speedup.

219
00:11:47,520 --> 00:11:49,804
AUDIENCE: Can you give an
example of an accelerated

220
00:11:49,804 --> 00:11:50,470
[? structure? ?]

221
00:11:50,470 --> 00:11:54,400
PRAMOOK KHUNGUM:
Basically, the most popular

222
00:11:54,400 --> 00:12:00,914
one is probably the--
what's it called?

223
00:12:00,914 --> 00:12:02,747
For example,
[? the KD tree, ?] for example.

224
00:12:02,747 --> 00:12:03,997
AUDIENCE: Oh, is it? [? OK. ?]

225
00:12:06,560 --> 00:12:09,790
PRAMOOK KHUNGUM: And
actually, I learned in class

226
00:12:09,790 --> 00:12:14,550
that basically, you can
achieve about 2x or 3x speedup

227
00:12:14,550 --> 00:12:21,310
by rearranging the data so that
you can trace four rays or 16

228
00:12:21,310 --> 00:12:24,360
rays at a time instead of doing
one at a time in the approach

229
00:12:24,360 --> 00:12:26,040
I'm taking.

230
00:12:26,040 --> 00:12:30,470
And if you do that, it's faster.

231
00:12:33,880 --> 00:12:38,850
In the program, I
have a lot of objects.

232
00:12:38,850 --> 00:12:41,390
And each of them has a
transform associated to them

233
00:12:41,390 --> 00:12:45,620
so that I can locate them
wherever I want and the object

234
00:12:45,620 --> 00:12:46,690
can be anything.

235
00:12:46,690 --> 00:12:48,940
But there's another
approach which

236
00:12:48,940 --> 00:12:55,580
says I just take anything to
be a collection of triangles.

237
00:12:55,580 --> 00:12:58,360
And if you have
triangles, the description

238
00:12:58,360 --> 00:13:00,450
is just three points
in space and you

239
00:13:00,450 --> 00:13:04,700
don't need transform to
do location or something

240
00:13:04,700 --> 00:13:05,380
like that.

241
00:13:05,380 --> 00:13:10,690
And basically, checking whether
a ray hits a triangle or not

242
00:13:10,690 --> 00:13:13,760
will be faster.

243
00:13:13,760 --> 00:13:16,443
So thank you very much.

244
00:13:16,443 --> 00:13:17,776
PROFESSOR: A few more questions?

245
00:13:21,046 --> 00:13:22,545
AUDIENCE: Do you
know that work that

246
00:13:22,545 --> 00:13:25,994
was done here by Julie
Dorsey in illumination?

247
00:13:25,994 --> 00:13:27,410
PRAMOOK KHUNGUM:
I have [? not. ?]

248
00:13:27,410 --> 00:13:31,150
AUDIENCE: She basically used
the trick of pre-computing,

249
00:13:31,150 --> 00:13:35,100
doing the ray tracing for
some fixed number of source

250
00:13:35,100 --> 00:13:39,990
locations and then just
rendering, in real time,

251
00:13:39,990 --> 00:13:46,110
by taking a linear superposition
of these ray-traced images.

252
00:13:46,110 --> 00:13:48,750
So long as you can
approximate your light source

253
00:13:48,750 --> 00:13:52,370
by a linear combination
of some fixed repertoire

254
00:13:52,370 --> 00:13:55,720
of light sources--
you didn't think

255
00:13:55,720 --> 00:13:59,904
about taking that
shortcut presumably.

256
00:13:59,904 --> 00:14:04,206
PRAMOOK KHUNGUM: I didn't think
about that shortcut basically.

257
00:14:04,206 --> 00:14:06,909
I [INAUDIBLE] know
something like that's--

258
00:14:06,909 --> 00:14:08,575
PROFESSOR: [INAUDIBLE]
few minutes left.

259
00:14:08,575 --> 00:14:10,440
AUDIENCE: Good.

260
00:14:10,440 --> 00:14:13,000
PRAMOOK KHUNGUM: Let's see.

261
00:14:13,000 --> 00:14:14,690
So [INAUDIBLE].

262
00:14:21,780 --> 00:14:24,071
AUDIENCE: So is it possible
to just connect it directly

263
00:14:24,071 --> 00:14:25,347
to a Playstation 3?

264
00:14:25,347 --> 00:14:26,930
PRAMOOK KHUNGUM: So
basically, there's

265
00:14:26,930 --> 00:14:30,730
this output.gga, which I'm
going to remove because I'm

266
00:14:30,730 --> 00:14:31,847
going to generate it.

267
00:14:31,847 --> 00:14:33,430
The scene I'm going
to render is going

268
00:14:33,430 --> 00:14:39,194
to be a Cornell Box with
a cube and a sphere in it.

269
00:14:39,194 --> 00:14:39,944
AUDIENCE: Pramook?

270
00:14:39,944 --> 00:14:40,819
PRAMOOK KHUNGUM: Yes?

271
00:14:40,819 --> 00:14:43,610
AUDIENCE: How easily could
you change the light sources?

272
00:14:43,610 --> 00:14:44,910
PRAMOOK KHUNGUM: I can
just move the object

273
00:14:44,910 --> 00:14:45,785
and the light source.

274
00:14:45,785 --> 00:14:48,250
AUDIENCE: OK, can you
do that [INAUDIBLE]?

275
00:14:48,250 --> 00:14:50,501
PRAMOOK KHUNGUM: Sure.

276
00:14:50,501 --> 00:14:52,250
Let's change it a point
light source then.

277
00:15:18,432 --> 00:15:20,480
[INAUDIBLE] relax.

278
00:15:20,480 --> 00:15:21,396
AUDIENCE: [INAUDIBLE].

279
00:15:26,319 --> 00:15:28,110
[? PROFESSOR: It just ?]
might be too slow.

280
00:15:28,110 --> 00:15:29,868
PRAMOOK KHUNGUM: I see.

281
00:15:29,868 --> 00:15:31,330
Here we go.

282
00:15:31,330 --> 00:15:34,730
So right now, the current
light source is this square.

283
00:15:34,730 --> 00:15:43,130
I'm going to comment it out and
change it to the point light

284
00:15:43,130 --> 00:15:43,962
source here.

285
00:15:47,898 --> 00:15:49,866
And then here we go.

286
00:15:55,759 --> 00:16:20,370
That's too quick, but--
basically, there's

287
00:16:20,370 --> 00:16:22,310
a point light source
located around here.

288
00:16:22,310 --> 00:16:25,050
So the scene is lit
with one point light.

289
00:16:25,050 --> 00:16:27,820
So as to expected,
the shadow is hard.

290
00:16:32,824 --> 00:16:33,990
AUDIENCE: One more question.

291
00:16:33,990 --> 00:16:34,430
PROFESSOR: One more question.

292
00:16:34,430 --> 00:16:35,550
PRAMOOK KHUNGUM: Sure.

293
00:16:35,550 --> 00:16:37,680
AUDIENCE: Just dividing
the scene equally

294
00:16:37,680 --> 00:16:39,360
seems like it give
good load balancing.

295
00:16:39,360 --> 00:16:41,764
If you had like roughly
equal complexity

296
00:16:41,764 --> 00:16:43,180
across different
sections-- if you

297
00:16:43,180 --> 00:16:46,920
have a really complex object
or if you have reflections that

298
00:16:46,920 --> 00:16:49,540
go much deeper in
some places in others,

299
00:16:49,540 --> 00:16:52,330
couldn't an equal division
lead to bad load balancing?

300
00:16:52,330 --> 00:16:54,486
Did you think about
other divisions?

301
00:16:54,486 --> 00:16:55,610
PRAMOOK KHUNGUM: I suppose.

302
00:16:55,610 --> 00:16:58,090
But basically, other
approach that I think of,

303
00:16:58,090 --> 00:17:02,880
which basically probably
you can divide the scene

304
00:17:02,880 --> 00:17:04,386
in a really fine-grain manner.

305
00:17:04,386 --> 00:17:06,511
AUDIENCE: Oh, you just do
like [? that striping. ?]

306
00:17:06,511 --> 00:17:07,386
PRAMOOK KHUNGUM: Yes.

307
00:17:07,386 --> 00:17:08,339
AUDIENCE: Yeah, OK.

308
00:17:08,339 --> 00:17:10,748
So is there a reason that
you chose [INAUDIBLE]?

309
00:17:10,748 --> 00:17:11,226
PROFESSOR SAMAN:
The communication

310
00:17:11,226 --> 00:17:13,616
how hard was it to get
this communication working?

311
00:17:13,616 --> 00:17:15,050
So what [? you've ?]
communicated,

312
00:17:15,050 --> 00:17:16,484
how often do you communicate?

313
00:17:19,352 --> 00:17:20,790
PROFESSOR: Repeat the question.

314
00:17:20,790 --> 00:17:25,170
PRAMOOK KHUNGUM:
Professor Saman asked

315
00:17:25,170 --> 00:17:29,030
how often does the SPU
communicate with one another,

316
00:17:29,030 --> 00:17:33,028
and the PPU, and how
often do I do that?

317
00:17:33,028 --> 00:17:34,510
PROFESSOR SAMAN: Yes.

318
00:17:34,510 --> 00:17:37,340
PRAMOOK KHUNGUM:
Basically, well,

319
00:17:37,340 --> 00:17:41,200
before the SPU starts
running, the PPU

320
00:17:41,200 --> 00:17:43,320
sends it the scene description.

321
00:17:43,320 --> 00:17:47,930
And then it sends it
the task that says,

322
00:17:47,930 --> 00:17:51,140
I want you to render
this pixel to this pixel.

323
00:17:51,140 --> 00:17:52,770
And that's it.

324
00:17:52,770 --> 00:17:58,090
And then the SPU just streams
the pixels to the main memory.

325
00:17:58,090 --> 00:18:00,970
PROFESSOR SAMAN: OK, so
you do a static balance,

326
00:18:00,970 --> 00:18:02,900
static distribution
of work [? around? ?]

327
00:18:02,900 --> 00:18:03,230
PRAMOOK KHUNGUM: Yes.

328
00:18:03,230 --> 00:18:04,475
PROFESSOR SAMAN: And
then when they're done,

329
00:18:04,475 --> 00:18:06,720
everybody sends it back,
and then you synchronize?

330
00:18:06,720 --> 00:18:07,600
PRAMOOK KHUNGUM: Yes.

331
00:18:07,600 --> 00:18:08,433
PROFESSOR SAMAN: OK.

332
00:18:08,433 --> 00:18:10,720
PRAMOOK KHUNGUM: Basically.

333
00:18:10,720 --> 00:18:12,520
PROFESSOR: OK, let's
thank the speaker.

334
00:18:12,520 --> 00:18:14,970
[APPLAUSE]