Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:00,000 --> 00:00:03,290
it all started over 50 years ago with a
2
00:00:03,290 --> 00:00:03,300
it all started over 50 years ago with a
3
00:00:03,300 --> 00:00:04,080
it all started over 50 years ago with a
simple question
4
00:00:04,080 --> 00:00:04,090
simple question
5
00:00:04,090 --> 00:00:06,050
simple question
[Music]
6
00:00:06,050 --> 00:00:06,060
[Music]
7
00:00:06,060 --> 00:00:09,710
[Music]
what if a computer could make pictures
8
00:00:09,710 --> 00:00:09,720
what if a computer could make pictures
9
00:00:09,720 --> 00:00:10,990
what if a computer could make pictures
how could we use them
10
00:00:10,990 --> 00:00:11,000
how could we use them
11
00:00:11,000 --> 00:00:12,650
how could we use them
[Music]
12
00:00:12,650 --> 00:00:12,660
[Music]
13
00:00:12,660 --> 00:00:13,990
[Music]
and what would they look like
14
00:00:13,990 --> 00:00:14,000
and what would they look like
15
00:00:14,000 --> 00:00:16,609
and what would they look like
Generations from now
16
00:00:16,609 --> 00:00:16,619
Generations from now
17
00:00:16,619 --> 00:00:25,970
Generations from now
hello folks I'm Mr computer image
18
00:00:25,970 --> 00:00:25,980
19
00:00:25,980 --> 00:00:34,729
[Music]
20
00:00:34,729 --> 00:00:34,739
[Music]
21
00:00:34,739 --> 00:00:43,920
[Music]
thank you
22
00:00:43,920 --> 00:00:43,930
23
00:00:43,930 --> 00:00:50,010
[Music]
24
00:00:50,010 --> 00:00:50,020
25
00:00:50,020 --> 00:00:52,460
[Music]
26
00:00:52,460 --> 00:00:52,470
[Music]
27
00:00:52,470 --> 00:00:53,010
[Music]
[Applause]
28
00:00:53,010 --> 00:00:53,020
[Applause]
29
00:00:53,020 --> 00:01:15,460
[Applause]
[Music]
30
00:01:15,460 --> 00:01:15,470
31
00:01:15,470 --> 00:01:23,870
[Music]
32
00:01:23,870 --> 00:01:23,880
[Music]
33
00:01:23,880 --> 00:01:36,770
[Music]
foreign
34
00:01:36,770 --> 00:01:36,780
35
00:01:36,780 --> 00:01:46,120
welcome in I hope you're hungry
36
00:01:46,120 --> 00:01:46,130
37
00:01:46,130 --> 00:02:00,590
[Music]
38
00:02:00,590 --> 00:02:00,600
39
00:02:00,600 --> 00:02:06,590
[Music]
40
00:02:06,590 --> 00:02:06,600
[Music]
41
00:02:06,600 --> 00:02:07,910
[Music]
this
42
00:02:07,910 --> 00:02:07,920
this
43
00:02:07,920 --> 00:02:10,910
this
was the Showcase demo
44
00:02:10,910 --> 00:02:10,920
was the Showcase demo
45
00:02:10,920 --> 00:02:12,350
was the Showcase demo
it had
46
00:02:12,350 --> 00:02:12,360
it had
47
00:02:12,360 --> 00:02:15,229
it had
two and a half million polygons or so
48
00:02:15,229 --> 00:02:15,239
two and a half million polygons or so
49
00:02:15,239 --> 00:02:17,210
two and a half million polygons or so
two rays per pixel
50
00:02:17,210 --> 00:02:17,220
two rays per pixel
51
00:02:17,220 --> 00:02:19,729
two rays per pixel
a couple of bounces per Ray
52
00:02:19,729 --> 00:02:19,739
a couple of bounces per Ray
53
00:02:19,739 --> 00:02:21,770
a couple of bounces per Ray
the demonstration was frankly at the
54
00:02:21,770 --> 00:02:21,780
the demonstration was frankly at the
55
00:02:21,780 --> 00:02:24,470
the demonstration was frankly at the
time incredibly beautiful
56
00:02:24,470 --> 00:02:24,480
time incredibly beautiful
57
00:02:24,480 --> 00:02:26,990
time incredibly beautiful
that was five years ago
58
00:02:26,990 --> 00:02:27,000
that was five years ago
59
00:02:27,000 --> 00:02:28,910
that was five years ago
now five years later
60
00:02:28,910 --> 00:02:28,920
now five years later
61
00:02:28,920 --> 00:02:31,089
now five years later
racer RTX
62
00:02:31,089 --> 00:02:31,099
racer RTX
63
00:02:31,099 --> 00:02:35,210
racer RTX
250 million polygons 100 times more
64
00:02:35,210 --> 00:02:35,220
250 million polygons 100 times more
65
00:02:35,220 --> 00:02:36,350
250 million polygons 100 times more
geometry
66
00:02:36,350 --> 00:02:36,360
geometry
67
00:02:36,360 --> 00:02:39,410
geometry
10 Rays per pixel about 10 balances per
68
00:02:39,410 --> 00:02:39,420
10 Rays per pixel about 10 balances per
69
00:02:39,420 --> 00:02:40,309
10 Rays per pixel about 10 balances per
Ray
70
00:02:40,309 --> 00:02:40,319
Ray
71
00:02:40,319 --> 00:02:43,550
Ray
and using dlss and for like seven out of
72
00:02:43,550 --> 00:02:43,560
and using dlss and for like seven out of
73
00:02:43,560 --> 00:02:45,170
and using dlss and for like seven out of
eight pixels
74
00:02:45,170 --> 00:02:45,180
eight pixels
75
00:02:45,180 --> 00:02:47,809
eight pixels
Computing only one out of eight and as a
76
00:02:47,809 --> 00:02:47,819
Computing only one out of eight and as a
77
00:02:47,819 --> 00:02:49,910
Computing only one out of eight and as a
result we're able to render this at 4K
78
00:02:49,910 --> 00:02:49,920
result we're able to render this at 4K
79
00:02:49,920 --> 00:03:08,210
result we're able to render this at 4K
scale it up to 4K 30 Hertz hit it
80
00:03:08,210 --> 00:03:08,220
81
00:03:08,220 --> 00:03:10,729
not bad for real time
82
00:03:10,729 --> 00:03:10,739
not bad for real time
83
00:03:10,739 --> 00:03:16,430
not bad for real time
thank you
84
00:03:16,430 --> 00:03:16,440
85
00:03:16,440 --> 00:03:19,250
while we were Reinventing computer
86
00:03:19,250 --> 00:03:19,260
while we were Reinventing computer
87
00:03:19,260 --> 00:03:23,089
while we were Reinventing computer
graphics with artificial intelligence
88
00:03:23,089 --> 00:03:23,099
graphics with artificial intelligence
89
00:03:23,099 --> 00:03:26,330
graphics with artificial intelligence
we were inventing the GPU
90
00:03:26,330 --> 00:03:26,340
we were inventing the GPU
91
00:03:26,340 --> 00:03:29,750
we were inventing the GPU
all together for artificial intelligence
92
00:03:29,750 --> 00:03:29,760
all together for artificial intelligence
93
00:03:29,760 --> 00:03:33,229
all together for artificial intelligence
the GPU when I came to see you last time
94
00:03:33,229 --> 00:03:33,239
the GPU when I came to see you last time
95
00:03:33,239 --> 00:03:35,390
the GPU when I came to see you last time
five years ago most people would say
96
00:03:35,390 --> 00:03:35,400
five years ago most people would say
97
00:03:35,400 --> 00:03:37,130
five years ago most people would say
that this is what a GPU looks like and
98
00:03:37,130 --> 00:03:37,140
that this is what a GPU looks like and
99
00:03:37,140 --> 00:03:39,290
that this is what a GPU looks like and
in fact this is the GPU that we
100
00:03:39,290 --> 00:03:39,300
in fact this is the GPU that we
101
00:03:39,300 --> 00:03:41,149
in fact this is the GPU that we
announced this is the touring GPU but
102
00:03:41,149 --> 00:03:41,159
announced this is the touring GPU but
103
00:03:41,159 --> 00:03:43,309
announced this is the touring GPU but
this is what a GPU is today
104
00:03:43,309 --> 00:03:43,319
this is what a GPU is today
105
00:03:43,319 --> 00:03:45,229
this is what a GPU is today
this GPU
106
00:03:45,229 --> 00:03:45,239
this GPU
107
00:03:45,239 --> 00:03:47,869
this GPU
is altogether one
108
00:03:47,869 --> 00:03:47,879
is altogether one
109
00:03:47,879 --> 00:03:50,690
is altogether one
trillion transistors
110
00:03:50,690 --> 00:03:50,700
trillion transistors
111
00:03:50,700 --> 00:03:54,229
trillion transistors
this GPU has 35 000 parts
112
00:03:54,229 --> 00:03:54,239
this GPU has 35 000 parts
113
00:03:54,239 --> 00:03:57,050
this GPU has 35 000 parts
it's manufactured by a robot like an
114
00:03:57,050 --> 00:03:57,060
it's manufactured by a robot like an
115
00:03:57,060 --> 00:03:58,309
it's manufactured by a robot like an
electric car
116
00:03:58,309 --> 00:03:58,319
electric car
117
00:03:58,319 --> 00:04:00,890
electric car
it weighs 70 pounds
118
00:04:00,890 --> 00:04:00,900
it weighs 70 pounds
119
00:04:00,900 --> 00:04:04,610
it weighs 70 pounds
consumes 6000 Watts
120
00:04:04,610 --> 00:04:04,620
consumes 6000 Watts
121
00:04:04,620 --> 00:04:06,890
consumes 6000 Watts
and this GPU revolutionized computer
122
00:04:06,890 --> 00:04:06,900
and this GPU revolutionized computer
123
00:04:06,900 --> 00:04:08,570
and this GPU revolutionized computer
science altogether
124
00:04:08,570 --> 00:04:08,580
science altogether
125
00:04:08,580 --> 00:04:11,089
science altogether
and 12 years later after 12 years
126
00:04:11,089 --> 00:04:11,099
and 12 years later after 12 years
127
00:04:11,099 --> 00:04:12,610
and 12 years later after 12 years
working on artificial intelligence
128
00:04:12,610 --> 00:04:12,620
working on artificial intelligence
129
00:04:12,620 --> 00:04:16,129
working on artificial intelligence
something gigantic happened degenerative
130
00:04:16,129 --> 00:04:16,139
something gigantic happened degenerative
131
00:04:16,139 --> 00:04:19,670
something gigantic happened degenerative
AI era is upon us
132
00:04:19,670 --> 00:04:19,680
AI era is upon us
133
00:04:19,680 --> 00:04:21,890
AI era is upon us
the iPhone moment of AI if you will
134
00:04:21,890 --> 00:04:21,900
the iPhone moment of AI if you will
135
00:04:21,900 --> 00:04:23,450
the iPhone moment of AI if you will
where all of the Technologies of
136
00:04:23,450 --> 00:04:23,460
where all of the Technologies of
137
00:04:23,460 --> 00:04:24,890
where all of the Technologies of
artificial intelligence came together in
138
00:04:24,890 --> 00:04:24,900
artificial intelligence came together in
139
00:04:24,900 --> 00:04:27,050
artificial intelligence came together in
such a way that it is now possible for
140
00:04:27,050 --> 00:04:27,060
such a way that it is now possible for
141
00:04:27,060 --> 00:04:28,850
such a way that it is now possible for
us to enjoy AI
142
00:04:28,850 --> 00:04:28,860
us to enjoy AI
143
00:04:28,860 --> 00:04:30,950
us to enjoy AI
in so many different applications
144
00:04:30,950 --> 00:04:30,960
in so many different applications
145
00:04:30,960 --> 00:04:32,629
in so many different applications
thousands of papers in just the last
146
00:04:32,629 --> 00:04:32,639
thousands of papers in just the last
147
00:04:32,639 --> 00:04:34,010
thousands of papers in just the last
several years
148
00:04:34,010 --> 00:04:34,020
several years
149
00:04:34,020 --> 00:04:35,629
several years
have been written about this area of
150
00:04:35,629 --> 00:04:35,639
have been written about this area of
151
00:04:35,639 --> 00:04:37,189
have been written about this area of
large language models and generative AI
152
00:04:37,189 --> 00:04:37,199
large language models and generative AI
153
00:04:37,199 --> 00:04:39,050
large language models and generative AI
here are some examples of some amazing
154
00:04:39,050 --> 00:04:39,060
here are some examples of some amazing
155
00:04:39,060 --> 00:04:41,689
here are some examples of some amazing
things this is the Adobe Firefly Adobe
156
00:04:41,689 --> 00:04:41,699
things this is the Adobe Firefly Adobe
157
00:04:41,699 --> 00:04:45,469
things this is the Adobe Firefly Adobe
Firefly does owl painting imagine
158
00:04:45,469 --> 00:04:45,479
Firefly does owl painting imagine
159
00:04:45,479 --> 00:04:48,230
Firefly does owl painting imagine
the space around the image that we never
160
00:04:48,230 --> 00:04:48,240
the space around the image that we never
161
00:04:48,240 --> 00:04:51,170
the space around the image that we never
captured move AI does mocap from just
162
00:04:51,170 --> 00:04:51,180
captured move AI does mocap from just
163
00:04:51,180 --> 00:04:52,430
captured move AI does mocap from just
video
164
00:04:52,430 --> 00:04:52,440
video
165
00:04:52,440 --> 00:04:54,650
video
this is on the upper right
166
00:04:54,650 --> 00:04:54,660
this is on the upper right
167
00:04:54,660 --> 00:04:59,930
this is on the upper right
you could you decide which one's real
168
00:04:59,930 --> 00:04:59,940
169
00:04:59,940 --> 00:05:02,930
I'm going with the left
170
00:05:02,930 --> 00:05:02,940
I'm going with the left
171
00:05:02,940 --> 00:05:05,150
I'm going with the left
viscom does sketch to image Guided by
172
00:05:05,150 --> 00:05:05,160
viscom does sketch to image Guided by
173
00:05:05,160 --> 00:05:07,129
viscom does sketch to image Guided by
language prompt this one's really cool
174
00:05:07,129 --> 00:05:07,139
language prompt this one's really cool
175
00:05:07,139 --> 00:05:08,510
language prompt this one's really cool
there are a lot of people who know how
176
00:05:08,510 --> 00:05:08,520
there are a lot of people who know how
177
00:05:08,520 --> 00:05:11,270
there are a lot of people who know how
to sketch and from the sketch and some
178
00:05:11,270 --> 00:05:11,280
to sketch and from the sketch and some
179
00:05:11,280 --> 00:05:13,249
to sketch and from the sketch and some
guidance from your language you could
180
00:05:13,249 --> 00:05:13,259
guidance from your language you could
181
00:05:13,259 --> 00:05:15,170
guidance from your language you could
generate something photorealistic and
182
00:05:15,170 --> 00:05:15,180
generate something photorealistic and
183
00:05:15,180 --> 00:05:16,189
generate something photorealistic and
rendered
184
00:05:16,189 --> 00:05:16,199
rendered
185
00:05:16,199 --> 00:05:17,689
rendered
the future of computer Graphics is
186
00:05:17,689 --> 00:05:17,699
the future of computer Graphics is
187
00:05:17,699 --> 00:05:19,670
the future of computer Graphics is
clearly going to be revolutionized and
188
00:05:19,670 --> 00:05:19,680
clearly going to be revolutionized and
189
00:05:19,680 --> 00:05:21,409
clearly going to be revolutionized and
this is really cool Wonder Dynamics not
190
00:05:21,409 --> 00:05:21,419
this is really cool Wonder Dynamics not
191
00:05:21,419 --> 00:05:22,730
this is really cool Wonder Dynamics not
not only is the name of the company call
192
00:05:22,730 --> 00:05:22,740
not only is the name of the company call
193
00:05:22,740 --> 00:05:24,950
not only is the name of the company call
but they do pose and lighting detection
194
00:05:24,950 --> 00:05:24,960
but they do pose and lighting detection
195
00:05:24,960 --> 00:05:28,129
but they do pose and lighting detection
and replace the actor with a CG
196
00:05:28,129 --> 00:05:28,139
and replace the actor with a CG
197
00:05:28,139 --> 00:05:29,749
and replace the actor with a CG
character
198
00:05:29,749 --> 00:05:29,759
character
199
00:05:29,759 --> 00:05:31,550
character
they're just it just goes on and on and
200
00:05:31,550 --> 00:05:31,560
they're just it just goes on and on and
201
00:05:31,560 --> 00:05:34,670
they're just it just goes on and on and
on the generative AI era has arrived
202
00:05:34,670 --> 00:05:34,680
on the generative AI era has arrived
203
00:05:34,680 --> 00:05:36,350
on the generative AI era has arrived
well
204
00:05:36,350 --> 00:05:36,360
well
205
00:05:36,360 --> 00:05:39,290
well
what's really profound though is that
206
00:05:39,290 --> 00:05:39,300
what's really profound though is that
207
00:05:39,300 --> 00:05:40,969
what's really profound though is that
when you take a step back and ask
208
00:05:40,969 --> 00:05:40,979
when you take a step back and ask
209
00:05:40,979 --> 00:05:42,170
when you take a step back and ask
yourself what is the meaning of
210
00:05:42,170 --> 00:05:42,180
yourself what is the meaning of
211
00:05:42,180 --> 00:05:43,730
yourself what is the meaning of
generative AI why is this such a big
212
00:05:43,730 --> 00:05:43,740
generative AI why is this such a big
213
00:05:43,740 --> 00:05:45,830
generative AI why is this such a big
deal why is it changing everything
214
00:05:45,830 --> 00:05:45,840
deal why is it changing everything
215
00:05:45,840 --> 00:05:47,689
deal why is it changing everything
well the reason for that is
216
00:05:47,689 --> 00:05:47,699
well the reason for that is
217
00:05:47,699 --> 00:05:49,070
well the reason for that is
first
218
00:05:49,070 --> 00:05:49,080
first
219
00:05:49,080 --> 00:05:52,070
first
human is the new programming language
220
00:05:52,070 --> 00:05:52,080
human is the new programming language
221
00:05:52,080 --> 00:05:53,950
human is the new programming language
we've democratized computer science
222
00:05:53,950 --> 00:05:53,960
we've democratized computer science
223
00:05:53,960 --> 00:05:56,330
we've democratized computer science
everybody can be a programmer now
224
00:05:56,330 --> 00:05:56,340
everybody can be a programmer now
225
00:05:56,340 --> 00:05:58,610
everybody can be a programmer now
because human language natural language
226
00:05:58,610 --> 00:05:58,620
because human language natural language
227
00:05:58,620 --> 00:05:59,990
because human language natural language
is the
228
00:05:59,990 --> 00:06:00,000
is the
229
00:06:00,000 --> 00:06:02,270
is the
best programming language and it's the
230
00:06:02,270 --> 00:06:02,280
best programming language and it's the
231
00:06:02,280 --> 00:06:03,830
best programming language and it's the
reason why chat GPT has been so popular
232
00:06:03,830 --> 00:06:03,840
reason why chat GPT has been so popular
233
00:06:03,840 --> 00:06:05,810
reason why chat GPT has been so popular
everybody can program that computer
234
00:06:05,810 --> 00:06:05,820
everybody can program that computer
235
00:06:05,820 --> 00:06:08,629
everybody can program that computer
large language model is a new Computing
236
00:06:08,629 --> 00:06:08,639
large language model is a new Computing
237
00:06:08,639 --> 00:06:09,590
large language model is a new Computing
platform
238
00:06:09,590 --> 00:06:09,600
platform
239
00:06:09,600 --> 00:06:11,270
platform
because now the programming language is
240
00:06:11,270 --> 00:06:11,280
because now the programming language is
241
00:06:11,280 --> 00:06:13,129
because now the programming language is
human and what your program the computer
242
00:06:13,129 --> 00:06:13,139
human and what your program the computer
243
00:06:13,139 --> 00:06:15,230
human and what your program the computer
understands larger language models and
244
00:06:15,230 --> 00:06:15,240
understands larger language models and
245
00:06:15,240 --> 00:06:17,629
understands larger language models and
generative AI is the new killer app
246
00:06:17,629 --> 00:06:17,639
generative AI is the new killer app
247
00:06:17,639 --> 00:06:20,749
generative AI is the new killer app
these three insights has gotten
248
00:06:20,749 --> 00:06:20,759
these three insights has gotten
249
00:06:20,759 --> 00:06:23,150
these three insights has gotten
everybody just insanely excited and
250
00:06:23,150 --> 00:06:23,160
everybody just insanely excited and
251
00:06:23,160 --> 00:06:25,070
everybody just insanely excited and
because for the very first time after 15
252
00:06:25,070 --> 00:06:25,080
because for the very first time after 15
253
00:06:25,080 --> 00:06:26,390
because for the very first time after 15
years or so
254
00:06:26,390 --> 00:06:26,400
years or so
255
00:06:26,400 --> 00:06:28,309
years or so
a new Computing platform has emerged
256
00:06:28,309 --> 00:06:28,319
a new Computing platform has emerged
257
00:06:28,319 --> 00:06:30,050
a new Computing platform has emerged
like the PC like the internet like
258
00:06:30,050 --> 00:06:30,060
like the PC like the internet like
259
00:06:30,060 --> 00:06:32,210
like the PC like the internet like
mobile cloud computing a new Computing
260
00:06:32,210 --> 00:06:32,220
mobile cloud computing a new Computing
261
00:06:32,220 --> 00:06:33,710
mobile cloud computing a new Computing
platform has merged and this new
262
00:06:33,710 --> 00:06:33,720
platform has merged and this new
263
00:06:33,720 --> 00:06:34,909
platform has merged and this new
Computing platform is going to enable
264
00:06:34,909 --> 00:06:34,919
Computing platform is going to enable
265
00:06:34,919 --> 00:06:36,770
Computing platform is going to enable
all kinds of new applications
266
00:06:36,770 --> 00:06:36,780
all kinds of new applications
267
00:06:36,780 --> 00:06:39,469
all kinds of new applications
but very differently than the past
268
00:06:39,469 --> 00:06:39,479
but very differently than the past
269
00:06:39,479 --> 00:06:42,110
but very differently than the past
this new Computing platform benefits
270
00:06:42,110 --> 00:06:42,120
this new Computing platform benefits
271
00:06:42,120 --> 00:06:44,930
this new Computing platform benefits
every single Computing platform before
272
00:06:44,930 --> 00:06:44,940
every single Computing platform before
273
00:06:44,940 --> 00:06:47,150
every single Computing platform before
it for the very first time this new
274
00:06:47,150 --> 00:06:47,160
it for the very first time this new
275
00:06:47,160 --> 00:06:49,010
it for the very first time this new
Computing platform not only enables new
276
00:06:49,010 --> 00:06:49,020
Computing platform not only enables new
277
00:06:49,020 --> 00:06:51,230
Computing platform not only enables new
applications in this new era but helps
278
00:06:51,230 --> 00:06:51,240
applications in this new era but helps
279
00:06:51,240 --> 00:06:53,689
applications in this new era but helps
every application in the old era
280
00:06:53,689 --> 00:06:53,699
every application in the old era
281
00:06:53,699 --> 00:06:55,790
every application in the old era
this is the reason why the industry is
282
00:06:55,790 --> 00:06:55,800
this is the reason why the industry is
283
00:06:55,800 --> 00:06:59,930
this is the reason why the industry is
moving so fast but one particular area
284
00:06:59,930 --> 00:06:59,940
moving so fast but one particular area
285
00:06:59,940 --> 00:07:02,629
moving so fast but one particular area
is extremely important which is the
286
00:07:02,629 --> 00:07:02,639
is extremely important which is the
287
00:07:02,639 --> 00:07:05,150
is extremely important which is the
basic scale out of the cloud
288
00:07:05,150 --> 00:07:05,160
basic scale out of the cloud
289
00:07:05,160 --> 00:07:06,529
basic scale out of the cloud
the basic skill out of the cloud
290
00:07:06,529 --> 00:07:06,539
the basic skill out of the cloud
291
00:07:06,539 --> 00:07:08,570
the basic skill out of the cloud
historically was based on
292
00:07:08,570 --> 00:07:08,580
historically was based on
293
00:07:08,580 --> 00:07:11,809
historically was based on
off-the-shelf CPUs x86 CPUs while
294
00:07:11,809 --> 00:07:11,819
off-the-shelf CPUs x86 CPUs while
295
00:07:11,819 --> 00:07:14,210
off-the-shelf CPUs x86 CPUs while
general purpose Computing is a horrible
296
00:07:14,210 --> 00:07:14,220
general purpose Computing is a horrible
297
00:07:14,220 --> 00:07:16,370
general purpose Computing is a horrible
way of doing generative Ai and so we
298
00:07:16,370 --> 00:07:16,380
way of doing generative Ai and so we
299
00:07:16,380 --> 00:07:18,770
way of doing generative Ai and so we
created a brand new processor for the
300
00:07:18,770 --> 00:07:18,780
created a brand new processor for the
301
00:07:18,780 --> 00:07:22,189
created a brand new processor for the
era of generative Ai and this is it this
302
00:07:22,189 --> 00:07:22,199
era of generative Ai and this is it this
303
00:07:22,199 --> 00:07:24,230
era of generative Ai and this is it this
is the Grace Hopper we announced Grace
304
00:07:24,230 --> 00:07:24,240
is the Grace Hopper we announced Grace
305
00:07:24,240 --> 00:07:25,969
is the Grace Hopper we announced Grace
Hopper in fact just only recently
306
00:07:25,969 --> 00:07:25,979
Hopper in fact just only recently
307
00:07:25,979 --> 00:07:28,790
Hopper in fact just only recently
several months ago and today we're
308
00:07:28,790 --> 00:07:28,800
several months ago and today we're
309
00:07:28,800 --> 00:07:30,290
several months ago and today we're
announcing that we're going to give it a
310
00:07:30,290 --> 00:07:30,300
announcing that we're going to give it a
311
00:07:30,300 --> 00:07:32,510
announcing that we're going to give it a
boost we're going to give this processor
312
00:07:32,510 --> 00:07:32,520
boost we're going to give this processor
313
00:07:32,520 --> 00:07:34,850
boost we're going to give this processor
a boost with the world's fastest memory
314
00:07:34,850 --> 00:07:34,860
a boost with the world's fastest memory
315
00:07:34,860 --> 00:07:37,249
a boost with the world's fastest memory
called hbm3e the world's fast memory
316
00:07:37,249 --> 00:07:37,259
called hbm3e the world's fast memory
317
00:07:37,259 --> 00:07:39,350
called hbm3e the world's fast memory
fastest memory now connected to Grace
318
00:07:39,350 --> 00:07:39,360
fastest memory now connected to Grace
319
00:07:39,360 --> 00:07:43,249
fastest memory now connected to Grace
Hopper we're calling it gh200
320
00:07:43,249 --> 00:07:43,259
Hopper we're calling it gh200
321
00:07:43,259 --> 00:07:46,129
Hopper we're calling it gh200
the chips are in production it has four
322
00:07:46,129 --> 00:07:46,139
the chips are in production it has four
323
00:07:46,139 --> 00:07:48,550
the chips are in production it has four
petaflops of Transformer engine
324
00:07:48,550 --> 00:07:48,560
petaflops of Transformer engine
325
00:07:48,560 --> 00:07:50,990
petaflops of Transformer engine
processing capability
326
00:07:50,990 --> 00:07:51,000
processing capability
327
00:07:51,000 --> 00:07:52,610
processing capability
and
328
00:07:52,610 --> 00:07:52,620
and
329
00:07:52,620 --> 00:07:56,270
and
now it has five terabytes per second of
330
00:07:56,270 --> 00:07:56,280
now it has five terabytes per second of
331
00:07:56,280 --> 00:07:59,510
now it has five terabytes per second of
HB and 3E performance so this is the new
332
00:07:59,510 --> 00:07:59,520
HB and 3E performance so this is the new
333
00:07:59,520 --> 00:08:01,210
HB and 3E performance so this is the new
G
334
00:08:01,210 --> 00:08:01,220
G
335
00:08:01,220 --> 00:08:04,189
G
h200 based on the architecture Grace
336
00:08:04,189 --> 00:08:04,199
h200 based on the architecture Grace
337
00:08:04,199 --> 00:08:06,050
h200 based on the architecture Grace
Hopper and a processor for this new
338
00:08:06,050 --> 00:08:06,060
Hopper and a processor for this new
339
00:08:06,060 --> 00:08:09,529
Hopper and a processor for this new
Computing era the CPU now has 144 cores
340
00:08:09,529 --> 00:08:09,539
Computing era the CPU now has 144 cores
341
00:08:09,539 --> 00:08:14,210
Computing era the CPU now has 144 cores
the GPU has 10 terabytes per second of
342
00:08:14,210 --> 00:08:14,220
the GPU has 10 terabytes per second of
343
00:08:14,220 --> 00:08:16,249
the GPU has 10 terabytes per second of
frame buffer bandwidth you could take
344
00:08:16,249 --> 00:08:16,259
frame buffer bandwidth you could take
345
00:08:16,259 --> 00:08:18,050
frame buffer bandwidth you could take
just about any large language model you
346
00:08:18,050 --> 00:08:18,060
just about any large language model you
347
00:08:18,060 --> 00:08:20,270
just about any large language model you
like and put it into this and it will
348
00:08:20,270 --> 00:08:20,280
like and put it into this and it will
349
00:08:20,280 --> 00:08:23,930
like and put it into this and it will
inference like crazy the inference cost
350
00:08:23,930 --> 00:08:23,940
inference like crazy the inference cost
351
00:08:23,940 --> 00:08:25,969
inference like crazy the inference cost
of large language models will drop
352
00:08:25,969 --> 00:08:25,979
of large language models will drop
353
00:08:25,979 --> 00:08:27,409
of large language models will drop
significantly because look how small
354
00:08:27,409 --> 00:08:27,419
significantly because look how small
355
00:08:27,419 --> 00:08:29,150
significantly because look how small
this computer is and you could scale
356
00:08:29,150 --> 00:08:29,160
this computer is and you could scale
357
00:08:29,160 --> 00:08:31,670
this computer is and you could scale
this out in the world's data centers you
358
00:08:31,670 --> 00:08:31,680
this out in the world's data centers you
359
00:08:31,680 --> 00:08:33,290
this out in the world's data centers you
can connect this with ethernet you can
360
00:08:33,290 --> 00:08:33,300
can connect this with ethernet you can
361
00:08:33,300 --> 00:08:35,029
can connect this with ethernet you can
connect it with the finiband and of
362
00:08:35,029 --> 00:08:35,039
connect it with the finiband and of
363
00:08:35,039 --> 00:08:36,709
connect it with the finiband and of
course there's all kinds of different
364
00:08:36,709 --> 00:08:36,719
course there's all kinds of different
365
00:08:36,719 --> 00:08:38,149
course there's all kinds of different
ways that you can scale it up this is
366
00:08:38,149 --> 00:08:38,159
ways that you can scale it up this is
367
00:08:38,159 --> 00:08:39,409
ways that you can scale it up this is
two gpus
368
00:08:39,409 --> 00:08:39,419
two gpus
369
00:08:39,419 --> 00:08:41,810
two gpus
but what if we would like to scale this
370
00:08:41,810 --> 00:08:41,820
but what if we would like to scale this
371
00:08:41,820 --> 00:08:44,870
but what if we would like to scale this
up into a much much larger GPU run it
372
00:08:44,870 --> 00:08:44,880
up into a much much larger GPU run it
373
00:08:44,880 --> 00:09:24,530
up into a much much larger GPU run it
please
374
00:09:24,530 --> 00:09:24,540
375
00:09:24,540 --> 00:09:50,920
foreign
376
00:09:50,920 --> 00:09:50,930
377
00:09:50,930 --> 00:09:56,449
[Applause]
378
00:09:56,449 --> 00:09:56,459
[Applause]
379
00:09:56,459 --> 00:10:00,350
[Applause]
this is actual size by the way
380
00:10:00,350 --> 00:10:00,360
this is actual size by the way
381
00:10:00,360 --> 00:10:04,370
this is actual size by the way
this is actual size and it probably even
382
00:10:04,370 --> 00:10:04,380
this is actual size and it probably even
383
00:10:04,380 --> 00:10:08,210
this is actual size and it probably even
runs crisis
384
00:10:08,210 --> 00:10:08,220
385
00:10:08,220 --> 00:10:10,790
the world's largest
386
00:10:10,790 --> 00:10:10,800
the world's largest
387
00:10:10,800 --> 00:10:15,650
the world's largest
single GPU one exoflops four petaflops
388
00:10:15,650 --> 00:10:15,660
single GPU one exoflops four petaflops
389
00:10:15,660 --> 00:10:19,070
single GPU one exoflops four petaflops
per grasshopper 256 connected by mv-link
390
00:10:19,070 --> 00:10:19,080
per grasshopper 256 connected by mv-link
391
00:10:19,080 --> 00:10:23,570
per grasshopper 256 connected by mv-link
into one giant system and so this is a
392
00:10:23,570 --> 00:10:23,580
into one giant system and so this is a
393
00:10:23,580 --> 00:10:25,970
into one giant system and so this is a
modern GPU so next time when you order a
394
00:10:25,970 --> 00:10:25,980
modern GPU so next time when you order a
395
00:10:25,980 --> 00:10:27,769
modern GPU so next time when you order a
GPU on Amazon don't be surprised if this
396
00:10:27,769 --> 00:10:27,779
GPU on Amazon don't be surprised if this
397
00:10:27,779 --> 00:10:31,269
GPU on Amazon don't be surprised if this
shows up
398
00:10:31,269 --> 00:10:31,279
399
00:10:31,279 --> 00:10:33,769
okay so that's how you take race Hopper
400
00:10:33,769 --> 00:10:33,779
okay so that's how you take race Hopper
401
00:10:33,779 --> 00:10:35,990
okay so that's how you take race Hopper
and scale it up into of course a giant
402
00:10:35,990 --> 00:10:36,000
and scale it up into of course a giant
403
00:10:36,000 --> 00:10:38,329
and scale it up into of course a giant
system future Frontier models will be
404
00:10:38,329 --> 00:10:38,339
system future Frontier models will be
405
00:10:38,339 --> 00:10:40,310
system future Frontier models will be
built this way and so let me show you
406
00:10:40,310 --> 00:10:40,320
built this way and so let me show you
407
00:10:40,320 --> 00:10:42,530
built this way and so let me show you
this this is how you would do it and so
408
00:10:42,530 --> 00:10:42,540
this this is how you would do it and so
409
00:10:42,540 --> 00:10:45,889
this this is how you would do it and so
now you would have a single Grace Hopper
410
00:10:45,889 --> 00:10:45,899
now you would have a single Grace Hopper
411
00:10:45,899 --> 00:10:48,050
now you would have a single Grace Hopper
in each one of these nodes this is the
412
00:10:48,050 --> 00:10:48,060
in each one of these nodes this is the
413
00:10:48,060 --> 00:10:49,550
in each one of these nodes this is the
way Computing was done in the past for
414
00:10:49,550 --> 00:10:49,560
way Computing was done in the past for
415
00:10:49,560 --> 00:10:51,290
way Computing was done in the past for
the last 60 years ever since the IBM
416
00:10:51,290 --> 00:10:51,300
the last 60 years ever since the IBM
417
00:10:51,300 --> 00:10:53,870
the last 60 years ever since the IBM
system 360 and for the last 60 years
418
00:10:53,870 --> 00:10:53,880
system 360 and for the last 60 years
419
00:10:53,880 --> 00:10:54,710
system 360 and for the last 60 years
that's the way we've been doing
420
00:10:54,710 --> 00:10:54,720
that's the way we've been doing
421
00:10:54,720 --> 00:10:57,230
that's the way we've been doing
Computing well now
422
00:10:57,230 --> 00:10:57,240
Computing well now
423
00:10:57,240 --> 00:10:59,030
Computing well now
general purpose of computing is going to
424
00:10:59,030 --> 00:10:59,040
general purpose of computing is going to
425
00:10:59,040 --> 00:11:01,009
general purpose of computing is going to
give way to accelerated Computing and AI
426
00:11:01,009 --> 00:11:01,019
give way to accelerated Computing and AI
427
00:11:01,019 --> 00:11:02,750
give way to accelerated Computing and AI
Computing every single application every
428
00:11:02,750 --> 00:11:02,760
Computing every single application every
429
00:11:02,760 --> 00:11:04,490
Computing every single application every
single database whenever you interact
430
00:11:04,490 --> 00:11:04,500
single database whenever you interact
431
00:11:04,500 --> 00:11:06,530
single database whenever you interact
with an app with a computer you'll
432
00:11:06,530 --> 00:11:06,540
with an app with a computer you'll
433
00:11:06,540 --> 00:11:07,910
with an app with a computer you'll
likely be first engaging a large
434
00:11:07,910 --> 00:11:07,920
likely be first engaging a large
435
00:11:07,920 --> 00:11:09,889
likely be first engaging a large
language model that large language model
436
00:11:09,889 --> 00:11:09,899
language model that large language model
437
00:11:09,899 --> 00:11:11,329
language model that large language model
will figure out what is your intention
438
00:11:11,329 --> 00:11:11,339
will figure out what is your intention
439
00:11:11,339 --> 00:11:13,250
will figure out what is your intention
what is your desire what are you trying
440
00:11:13,250 --> 00:11:13,260
what is your desire what are you trying
441
00:11:13,260 --> 00:11:15,650
what is your desire what are you trying
to do given the context and present the
442
00:11:15,650 --> 00:11:15,660
to do given the context and present the
443
00:11:15,660 --> 00:11:16,970
to do given the context and present the
information to you in the best possible
444
00:11:16,970 --> 00:11:16,980
information to you in the best possible
445
00:11:16,980 --> 00:11:20,329
information to you in the best possible
way well if you were to have an ISO
446
00:11:20,329 --> 00:11:20,339
way well if you were to have an ISO
447
00:11:20,339 --> 00:11:22,790
way well if you were to have an ISO
budget way of processing that workload
448
00:11:22,790 --> 00:11:22,800
budget way of processing that workload
449
00:11:22,800 --> 00:11:24,410
budget way of processing that workload
it would take let me just choose the
450
00:11:24,410 --> 00:11:24,420
it would take let me just choose the
451
00:11:24,420 --> 00:11:25,670
it would take let me just choose the
number 100 million dollars and 100
452
00:11:25,670 --> 00:11:25,680
number 100 million dollars and 100
453
00:11:25,680 --> 00:11:26,870
number 100 million dollars and 100
million dollars would be a reasonably
454
00:11:26,870 --> 00:11:26,880
million dollars would be a reasonably
455
00:11:26,880 --> 00:11:28,730
million dollars would be a reasonably
small data center these days 100 million
456
00:11:28,730 --> 00:11:28,740
small data center these days 100 million
457
00:11:28,740 --> 00:11:31,310
small data center these days 100 million
dollars will buy you about 8 800 x86
458
00:11:31,310 --> 00:11:31,320
dollars will buy you about 8 800 x86
459
00:11:31,320 --> 00:11:33,710
dollars will buy you about 8 800 x86
gpus and we can take about five
460
00:11:33,710 --> 00:11:33,720
gpus and we can take about five
461
00:11:33,720 --> 00:11:35,990
gpus and we can take about five
megawatts to operate that and I
462
00:11:35,990 --> 00:11:36,000
megawatts to operate that and I
463
00:11:36,000 --> 00:11:38,870
megawatts to operate that and I
normalize the performance into 1X
464
00:11:38,870 --> 00:11:38,880
normalize the performance into 1X
465
00:11:38,880 --> 00:11:40,970
normalize the performance into 1X
using the exact same budget with
466
00:11:40,970 --> 00:11:40,980
using the exact same budget with
467
00:11:40,980 --> 00:11:44,030
using the exact same budget with
accelerated Computing Grace Hopper
468
00:11:44,030 --> 00:11:44,040
accelerated Computing Grace Hopper
469
00:11:44,040 --> 00:11:46,190
accelerated Computing Grace Hopper
it would consume only three megawatts
470
00:11:46,190 --> 00:11:46,200
it would consume only three megawatts
471
00:11:46,200 --> 00:11:49,970
it would consume only three megawatts
but your throughput goes up by an order
472
00:11:49,970 --> 00:11:49,980
but your throughput goes up by an order
473
00:11:49,980 --> 00:11:51,470
but your throughput goes up by an order
of magnitude
474
00:11:51,470 --> 00:11:51,480
of magnitude
475
00:11:51,480 --> 00:11:54,050
of magnitude
basically the Energy Efficiency the cost
476
00:11:54,050 --> 00:11:54,060
basically the Energy Efficiency the cost
477
00:11:54,060 --> 00:11:57,410
basically the Energy Efficiency the cost
efficiency of accelerated Computing for
478
00:11:57,410 --> 00:11:57,420
efficiency of accelerated Computing for
479
00:11:57,420 --> 00:12:00,250
efficiency of accelerated Computing for
generative AI applications is about 20x
480
00:12:00,250 --> 00:12:00,260
generative AI applications is about 20x
481
00:12:00,260 --> 00:12:03,350
generative AI applications is about 20x
20x in Moore's Law and just the current
482
00:12:03,350 --> 00:12:03,360
20x in Moore's Law and just the current
483
00:12:03,360 --> 00:12:06,050
20x in Moore's Law and just the current
way of scaling CPUs that would be a very
484
00:12:06,050 --> 00:12:06,060
way of scaling CPUs that would be a very
485
00:12:06,060 --> 00:12:08,389
way of scaling CPUs that would be a very
very long time and so this is a giant
486
00:12:08,389 --> 00:12:08,399
very long time and so this is a giant
487
00:12:08,399 --> 00:12:11,449
very long time and so this is a giant
Step Up in efficiency and throughput so
488
00:12:11,449 --> 00:12:11,459
Step Up in efficiency and throughput so
489
00:12:11,459 --> 00:12:14,269
Step Up in efficiency and throughput so
this is ISO budget let's take a look at
490
00:12:14,269 --> 00:12:14,279
this is ISO budget let's take a look at
491
00:12:14,279 --> 00:12:16,790
this is ISO budget let's take a look at
this now again and let's go through ISO
492
00:12:16,790 --> 00:12:16,800
this now again and let's go through ISO
493
00:12:16,800 --> 00:12:19,130
this now again and let's go through ISO
workload suppose your intention was to
494
00:12:19,130 --> 00:12:19,140
workload suppose your intention was to
495
00:12:19,140 --> 00:12:21,650
workload suppose your intention was to
provide a service and that service has
496
00:12:21,650 --> 00:12:21,660
provide a service and that service has
497
00:12:21,660 --> 00:12:23,569
provide a service and that service has
so many number of users and so your
498
00:12:23,569 --> 00:12:23,579
so many number of users and so your
499
00:12:23,579 --> 00:12:25,130
so many number of users and so your
workload is fairly well understood plus
500
00:12:25,130 --> 00:12:25,140
workload is fairly well understood plus
501
00:12:25,140 --> 00:12:27,889
workload is fairly well understood plus
or minus and so with ISO workload this
502
00:12:27,889 --> 00:12:27,899
or minus and so with ISO workload this
503
00:12:27,899 --> 00:12:31,130
or minus and so with ISO workload this
1X 100 million dollars using general
504
00:12:31,130 --> 00:12:31,140
1X 100 million dollars using general
505
00:12:31,140 --> 00:12:32,690
1X 100 million dollars using general
purpose computing
506
00:12:32,690 --> 00:12:32,700
purpose computing
507
00:12:32,700 --> 00:12:34,490
purpose computing
and using accelerated Computing Grace
508
00:12:34,490 --> 00:12:34,500
and using accelerated Computing Grace
509
00:12:34,500 --> 00:12:35,449
and using accelerated Computing Grace
Hopper
510
00:12:35,449 --> 00:12:35,459
Hopper
511
00:12:35,459 --> 00:12:37,910
Hopper
it would only cost eight million dollars
512
00:12:37,910 --> 00:12:37,920
it would only cost eight million dollars
513
00:12:37,920 --> 00:12:41,329
it would only cost eight million dollars
eight million dollars and only 260
514
00:12:41,329 --> 00:12:41,339
eight million dollars and only 260
515
00:12:41,339 --> 00:12:46,430
eight million dollars and only 260
kilowatts so 20 times less power and 12
516
00:12:46,430 --> 00:12:46,440
kilowatts so 20 times less power and 12
517
00:12:46,440 --> 00:12:49,009
kilowatts so 20 times less power and 12
times less cost this is the reason why
518
00:12:49,009 --> 00:12:49,019
times less cost this is the reason why
519
00:12:49,019 --> 00:12:51,590
times less cost this is the reason why
accelerated Computing is going to be the
520
00:12:51,590 --> 00:12:51,600
accelerated Computing is going to be the
521
00:12:51,600 --> 00:12:53,810
accelerated Computing is going to be the
path forward and this is the reason why
522
00:12:53,810 --> 00:12:53,820
path forward and this is the reason why
523
00:12:53,820 --> 00:12:55,790
path forward and this is the reason why
the world's data centers are very
524
00:12:55,790 --> 00:12:55,800
the world's data centers are very
525
00:12:55,800 --> 00:12:58,009
the world's data centers are very
quickly transitioning to accelerated
526
00:12:58,009 --> 00:12:58,019
quickly transitioning to accelerated
527
00:12:58,019 --> 00:12:59,870
quickly transitioning to accelerated
Computing and some people say and you
528
00:12:59,870 --> 00:12:59,880
Computing and some people say and you
529
00:12:59,880 --> 00:13:01,250
Computing and some people say and you
guys might have heard I don't know who
530
00:13:01,250 --> 00:13:01,260
guys might have heard I don't know who
531
00:13:01,260 --> 00:13:03,889
guys might have heard I don't know who
said it but the more you buy
532
00:13:03,889 --> 00:13:03,899
said it but the more you buy
533
00:13:03,899 --> 00:13:06,350
said it but the more you buy
the the more you save and and that's
534
00:13:06,350 --> 00:13:06,360
the the more you save and and that's
535
00:13:06,360 --> 00:13:08,690
the the more you save and and that's
that's wisdom
536
00:13:08,690 --> 00:13:08,700
that's wisdom
537
00:13:08,700 --> 00:13:10,670
that's wisdom
let's talk about a couple new things and
538
00:13:10,670 --> 00:13:10,680
let's talk about a couple new things and
539
00:13:10,680 --> 00:13:12,710
let's talk about a couple new things and
so today I want to talk about Omniverse
540
00:13:12,710 --> 00:13:12,720
so today I want to talk about Omniverse
541
00:13:12,720 --> 00:13:14,269
so today I want to talk about Omniverse
and generative Ai and how they come
542
00:13:14,269 --> 00:13:14,279
and generative Ai and how they come
543
00:13:14,279 --> 00:13:16,190
and generative Ai and how they come
together the first thing that we already
544
00:13:16,190 --> 00:13:16,200
together the first thing that we already
545
00:13:16,200 --> 00:13:17,930
together the first thing that we already
established is that graphics and
546
00:13:17,930 --> 00:13:17,940
established is that graphics and
547
00:13:17,940 --> 00:13:20,930
established is that graphics and
artificial intelligence are inseparable
548
00:13:20,930 --> 00:13:20,940
artificial intelligence are inseparable
549
00:13:20,940 --> 00:13:25,190
artificial intelligence are inseparable
that Graphics needs Ai and AI needs
550
00:13:25,190 --> 00:13:25,200
that Graphics needs Ai and AI needs
551
00:13:25,200 --> 00:13:28,490
that Graphics needs Ai and AI needs
Graphics Graphics needs Ai and AI needs
552
00:13:28,490 --> 00:13:28,500
Graphics Graphics needs Ai and AI needs
553
00:13:28,500 --> 00:13:29,930
Graphics Graphics needs Ai and AI needs
graphics and so the thing that we could
554
00:13:29,930 --> 00:13:29,940
graphics and so the thing that we could
555
00:13:29,940 --> 00:13:32,329
graphics and so the thing that we could
do is we could create a virtual world
556
00:13:32,329 --> 00:13:32,339
do is we could create a virtual world
557
00:13:32,339 --> 00:13:34,910
do is we could create a virtual world
that is physically simulated physics
558
00:13:34,910 --> 00:13:34,920
that is physically simulated physics
559
00:13:34,920 --> 00:13:36,530
that is physically simulated physics
simulator that allows an artificial
560
00:13:36,530 --> 00:13:36,540
simulator that allows an artificial
561
00:13:36,540 --> 00:13:38,269
simulator that allows an artificial
intelligence to learn how to perceive
562
00:13:38,269 --> 00:13:38,279
intelligence to learn how to perceive
563
00:13:38,279 --> 00:13:39,530
intelligence to learn how to perceive
the environment
564
00:13:39,530 --> 00:13:39,540
the environment
565
00:13:39,540 --> 00:13:41,629
the environment
using a vision Transformer Maybe
566
00:13:41,629 --> 00:13:41,639
using a vision Transformer Maybe
567
00:13:41,639 --> 00:13:43,850
using a vision Transformer Maybe
and to use reinforcement learning to
568
00:13:43,850 --> 00:13:43,860
and to use reinforcement learning to
569
00:13:43,860 --> 00:13:46,910
and to use reinforcement learning to
understand the consequences of its
570
00:13:46,910 --> 00:13:46,920
understand the consequences of its
571
00:13:46,920 --> 00:13:48,650
understand the consequences of its
physical actions and learn how to
572
00:13:48,650 --> 00:13:48,660
physical actions and learn how to
573
00:13:48,660 --> 00:13:50,990
physical actions and learn how to
animate and learn how to articulate to
574
00:13:50,990 --> 00:13:51,000
animate and learn how to articulate to
575
00:13:51,000 --> 00:13:52,670
animate and learn how to articulate to
achieve a particular goal
576
00:13:52,670 --> 00:13:52,680
achieve a particular goal
577
00:13:52,680 --> 00:13:56,629
achieve a particular goal
and so one mission of a connected
578
00:13:56,629 --> 00:13:56,639
and so one mission of a connected
579
00:13:56,639 --> 00:13:59,210
and so one mission of a connected
artificial intelligence system and a
580
00:13:59,210 --> 00:13:59,220
artificial intelligence system and a
581
00:13:59,220 --> 00:14:00,290
artificial intelligence system and a
virtual world system that we call
582
00:14:00,290 --> 00:14:00,300
virtual world system that we call
583
00:14:00,300 --> 00:14:03,410
virtual world system that we call
Omniverse is so that the future of AI
584
00:14:03,410 --> 00:14:03,420
Omniverse is so that the future of AI
585
00:14:03,420 --> 00:14:06,050
Omniverse is so that the future of AI
could be physically grounded the number
586
00:14:06,050 --> 00:14:06,060
could be physically grounded the number
587
00:14:06,060 --> 00:14:08,810
could be physically grounded the number
of applications is really quite exciting
588
00:14:08,810 --> 00:14:08,820
of applications is really quite exciting
589
00:14:08,820 --> 00:14:10,970
of applications is really quite exciting
because as we know the largest
590
00:14:10,970 --> 00:14:10,980
because as we know the largest
591
00:14:10,980 --> 00:14:12,470
because as we know the largest
Industries in the world are heavy
592
00:14:12,470 --> 00:14:12,480
Industries in the world are heavy
593
00:14:12,480 --> 00:14:14,150
Industries in the world are heavy
industry and those heavy Industries are
594
00:14:14,150 --> 00:14:14,160
industry and those heavy Industries are
595
00:14:14,160 --> 00:14:16,430
industry and those heavy Industries are
physics-based physically based and so
596
00:14:16,430 --> 00:14:16,440
physics-based physically based and so
597
00:14:16,440 --> 00:14:19,850
physics-based physically based and so
first application is so that AI
598
00:14:19,850 --> 00:14:19,860
first application is so that AI
599
00:14:19,860 --> 00:14:23,150
first application is so that AI
can learn in a virtual world
600
00:14:23,150 --> 00:14:23,160
can learn in a virtual world
601
00:14:23,160 --> 00:14:25,730
can learn in a virtual world
an application the second reason why AI
602
00:14:25,730 --> 00:14:25,740
an application the second reason why AI
603
00:14:25,740 --> 00:14:28,310
an application the second reason why AI
is and computer Graphics are inseparable
604
00:14:28,310 --> 00:14:28,320
is and computer Graphics are inseparable
605
00:14:28,320 --> 00:14:31,910
is and computer Graphics are inseparable
is that AI will help also to create
606
00:14:31,910 --> 00:14:31,920
is that AI will help also to create
607
00:14:31,920 --> 00:14:33,650
is that AI will help also to create
these Virtual Worlds
608
00:14:33,650 --> 00:14:33,660
these Virtual Worlds
609
00:14:33,660 --> 00:14:36,110
these Virtual Worlds
well let's take a look at what wpp the
610
00:14:36,110 --> 00:14:36,120
well let's take a look at what wpp the
611
00:14:36,120 --> 00:14:39,350
well let's take a look at what wpp the
world's largest Ad Agency and byd the
612
00:14:39,350 --> 00:14:39,360
world's largest Ad Agency and byd the
613
00:14:39,360 --> 00:14:40,970
world's largest Ad Agency and byd the
world's largest electric vehicle maker
614
00:14:40,970 --> 00:14:40,980
world's largest electric vehicle maker
615
00:14:40,980 --> 00:14:43,189
world's largest electric vehicle maker
how they're using Omniverse and
616
00:14:43,189 --> 00:14:43,199
how they're using Omniverse and
617
00:14:43,199 --> 00:14:45,110
how they're using Omniverse and
generative AI in their work play it
618
00:14:45,110 --> 00:14:45,120
generative AI in their work play it
619
00:14:45,120 --> 00:14:47,710
generative AI in their work play it
please
620
00:14:47,710 --> 00:14:47,720
621
00:14:47,720 --> 00:14:50,210
wpp is building the next generation of
622
00:14:50,210 --> 00:14:50,220
wpp is building the next generation of
623
00:14:50,220 --> 00:14:52,189
wpp is building the next generation of
car configurators for automotive giant
624
00:14:52,189 --> 00:14:52,199
car configurators for automotive giant
625
00:14:52,199 --> 00:14:54,949
car configurators for automotive giant
byds denza luxury brand powered by
626
00:14:54,949 --> 00:14:54,959
byds denza luxury brand powered by
627
00:14:54,959 --> 00:14:57,710
byds denza luxury brand powered by
Omniverse cloud and generative AI open
628
00:14:57,710 --> 00:14:57,720
Omniverse cloud and generative AI open
629
00:14:57,720 --> 00:15:00,829
Omniverse cloud and generative AI open
USD and Omniverse Cloud allows denza to
630
00:15:00,829 --> 00:15:00,839
USD and Omniverse Cloud allows denza to
631
00:15:00,839 --> 00:15:02,210
USD and Omniverse Cloud allows denza to
connect High Fidelity data from
632
00:15:02,210 --> 00:15:02,220
connect High Fidelity data from
633
00:15:02,220 --> 00:15:04,490
connect High Fidelity data from
industry-leading CAD tools to create a
634
00:15:04,490 --> 00:15:04,500
industry-leading CAD tools to create a
635
00:15:04,500 --> 00:15:06,350
industry-leading CAD tools to create a
physically accurate real-time digital
636
00:15:06,350 --> 00:15:06,360
physically accurate real-time digital
637
00:15:06,360 --> 00:15:08,449
physically accurate real-time digital
twin of its N7
638
00:15:08,449 --> 00:15:08,459
twin of its N7
639
00:15:08,459 --> 00:15:11,210
twin of its N7
wpp artists can work seamlessly on this
640
00:15:11,210 --> 00:15:11,220
wpp artists can work seamlessly on this
641
00:15:11,220 --> 00:15:12,949
wpp artists can work seamlessly on this
model in the same Omniverse Cloud
642
00:15:12,949 --> 00:15:12,959
model in the same Omniverse Cloud
643
00:15:12,959 --> 00:15:14,750
model in the same Omniverse Cloud
environment with their preferred tools
644
00:15:14,750 --> 00:15:14,760
environment with their preferred tools
645
00:15:14,760 --> 00:15:17,449
environment with their preferred tools
from Autodesk Adobe and side effects to
646
00:15:17,449 --> 00:15:17,459
from Autodesk Adobe and side effects to
647
00:15:17,459 --> 00:15:19,129
from Autodesk Adobe and side effects to
deliver the next era of automotive
648
00:15:19,129 --> 00:15:19,139
deliver the next era of automotive
649
00:15:19,139 --> 00:15:22,129
deliver the next era of automotive
digitalization and immersive experiences
650
00:15:22,129 --> 00:15:22,139
digitalization and immersive experiences
651
00:15:22,139 --> 00:15:24,410
digitalization and immersive experiences
today's configurators require hundreds
652
00:15:24,410 --> 00:15:24,420
today's configurators require hundreds
653
00:15:24,420 --> 00:15:25,670
today's configurators require hundreds
of thousands of images to be
654
00:15:25,670 --> 00:15:25,680
of thousands of images to be
655
00:15:25,680 --> 00:15:27,590
of thousands of images to be
pre-rendered to represent all possible
656
00:15:27,590 --> 00:15:27,600
pre-rendered to represent all possible
657
00:15:27,600 --> 00:15:29,449
pre-rendered to represent all possible
options and variants
658
00:15:29,449 --> 00:15:29,459
options and variants
659
00:15:29,459 --> 00:15:32,810
options and variants
open USD makes it possible for wpp to
660
00:15:32,810 --> 00:15:32,820
open USD makes it possible for wpp to
661
00:15:32,820 --> 00:15:34,370
open USD makes it possible for wpp to
create a super digital twin of the card
662
00:15:34,370 --> 00:15:34,380
create a super digital twin of the card
663
00:15:34,380 --> 00:15:36,470
create a super digital twin of the card
that includes all possible variants in
664
00:15:36,470 --> 00:15:36,480
that includes all possible variants in
665
00:15:36,480 --> 00:15:38,629
that includes all possible variants in
one single asset deployed as a fully
666
00:15:38,629 --> 00:15:38,639
one single asset deployed as a fully
667
00:15:38,639 --> 00:15:40,970
one single asset deployed as a fully
interactive 3D configurator on Omniverse
668
00:15:40,970 --> 00:15:40,980
interactive 3D configurator on Omniverse
669
00:15:40,980 --> 00:15:43,250
interactive 3D configurator on Omniverse
Cloud gdn a network that can stream High
670
00:15:43,250 --> 00:15:43,260
Cloud gdn a network that can stream High
671
00:15:43,260 --> 00:15:46,069
Cloud gdn a network that can stream High
Fidelity real-time 3D experiences to
672
00:15:46,069 --> 00:15:46,079
Fidelity real-time 3D experiences to
673
00:15:46,079 --> 00:15:48,769
Fidelity real-time 3D experiences to
devices in over 100 regions were used to
674
00:15:48,769 --> 00:15:48,779
devices in over 100 regions were used to
675
00:15:48,779 --> 00:15:50,629
devices in over 100 regions were used to
generate thousands of individual pieces
676
00:15:50,629 --> 00:15:50,639
generate thousands of individual pieces
677
00:15:50,639 --> 00:15:52,430
generate thousands of individual pieces
of content that comprise a global
678
00:15:52,430 --> 00:15:52,440
of content that comprise a global
679
00:15:52,440 --> 00:15:55,250
of content that comprise a global
marketing campaign the USD model is
680
00:15:55,250 --> 00:15:55,260
marketing campaign the USD model is
681
00:15:55,260 --> 00:15:57,350
marketing campaign the USD model is
placed in a 3D environment that can
682
00:15:57,350 --> 00:15:57,360
placed in a 3D environment that can
683
00:15:57,360 --> 00:15:58,790
placed in a 3D environment that can
either be scanned from The Real World
684
00:15:58,790 --> 00:15:58,800
either be scanned from The Real World
685
00:15:58,800 --> 00:16:01,490
either be scanned from The Real World
using lidar and virtual production were
686
00:16:01,490 --> 00:16:01,500
using lidar and virtual production were
687
00:16:01,500 --> 00:16:03,290
using lidar and virtual production were
created in seconds with generative AI
688
00:16:03,290 --> 00:16:03,300
created in seconds with generative AI
689
00:16:03,300 --> 00:16:05,449
created in seconds with generative AI
tools from organizations such as Adobe
690
00:16:05,449 --> 00:16:05,459
tools from organizations such as Adobe
691
00:16:05,459 --> 00:16:08,389
tools from organizations such as Adobe
and Shutterstock this Innovative wpp
692
00:16:08,389 --> 00:16:08,399
and Shutterstock this Innovative wpp
693
00:16:08,399 --> 00:16:10,730
and Shutterstock this Innovative wpp
solution for byd brings generative Ai
694
00:16:10,730 --> 00:16:10,740
solution for byd brings generative Ai
695
00:16:10,740 --> 00:16:12,230
solution for byd brings generative Ai
and collab rendered real-time 3D
696
00:16:12,230 --> 00:16:12,240
and collab rendered real-time 3D
697
00:16:12,240 --> 00:16:14,389
and collab rendered real-time 3D
together for the first time powering the
698
00:16:14,389 --> 00:16:14,399
together for the first time powering the
699
00:16:14,399 --> 00:16:17,509
together for the first time powering the
next generation of e-commerce
700
00:16:17,509 --> 00:16:17,519
next generation of e-commerce
701
00:16:17,519 --> 00:16:21,829
next generation of e-commerce
kind of love that
702
00:16:21,829 --> 00:16:21,839
703
00:16:21,839 --> 00:16:24,650
everything everything was rendered in
704
00:16:24,650 --> 00:16:24,660
everything everything was rendered in
705
00:16:24,660 --> 00:16:27,110
everything everything was rendered in
real time nothing was pre-rendered every
706
00:16:27,110 --> 00:16:27,120
real time nothing was pre-rendered every
707
00:16:27,120 --> 00:16:28,730
real time nothing was pre-rendered every
single every single scene that you saw
708
00:16:28,730 --> 00:16:28,740
single every single scene that you saw
709
00:16:28,740 --> 00:16:30,410
single every single scene that you saw
was rendered in real time nothing was
710
00:16:30,410 --> 00:16:30,420
was rendered in real time nothing was
711
00:16:30,420 --> 00:16:32,810
was rendered in real time nothing was
changed you literally take the cad you
712
00:16:32,810 --> 00:16:32,820
changed you literally take the cad you
713
00:16:32,820 --> 00:16:35,170
changed you literally take the cad you
drag it into Omniverse you tell an AI
714
00:16:35,170 --> 00:16:35,180
drag it into Omniverse you tell an AI
715
00:16:35,180 --> 00:16:37,850
drag it into Omniverse you tell an AI
synthesize and generate an environment
716
00:16:37,850 --> 00:16:37,860
synthesize and generate an environment
717
00:16:37,860 --> 00:16:40,310
synthesize and generate an environment
and all of a sudden the car up here is
718
00:16:40,310 --> 00:16:40,320
and all of a sudden the car up here is
719
00:16:40,320 --> 00:16:42,410
and all of a sudden the car up here is
wherever you like it to be so this is a
720
00:16:42,410 --> 00:16:42,420
wherever you like it to be so this is a
721
00:16:42,420 --> 00:16:46,129
wherever you like it to be so this is a
one example of how generative Ai and
722
00:16:46,129 --> 00:16:46,139
one example of how generative Ai and
723
00:16:46,139 --> 00:16:48,530
one example of how generative Ai and
human designs come together to create
724
00:16:48,530 --> 00:16:48,540
human designs come together to create
725
00:16:48,540 --> 00:16:51,170
human designs come together to create
these incredible applications and so how
726
00:16:51,170 --> 00:16:51,180
these incredible applications and so how
727
00:16:51,180 --> 00:16:53,449
these incredible applications and so how
do we do this applications generative AI
728
00:16:53,449 --> 00:16:53,459
do we do this applications generative AI
729
00:16:53,459 --> 00:16:55,430
do we do this applications generative AI
models are making tremendous
730
00:16:55,430 --> 00:16:55,440
models are making tremendous
731
00:16:55,440 --> 00:16:58,189
models are making tremendous
breakthroughs in fact the hopper GPU is
732
00:16:58,189 --> 00:16:58,199
breakthroughs in fact the hopper GPU is
733
00:16:58,199 --> 00:17:00,710
breakthroughs in fact the hopper GPU is
impossible to design by humans we needed
734
00:17:00,710 --> 00:17:00,720
impossible to design by humans we needed
735
00:17:00,720 --> 00:17:03,170
impossible to design by humans we needed
AIS and generative models to help us
736
00:17:03,170 --> 00:17:03,180
AIS and generative models to help us
737
00:17:03,180 --> 00:17:05,929
AIS and generative models to help us
find the way to design this thing in
738
00:17:05,929 --> 00:17:05,939
find the way to design this thing in
739
00:17:05,939 --> 00:17:08,030
find the way to design this thing in
such a high performance way and so it
740
00:17:08,030 --> 00:17:08,040
such a high performance way and so it
741
00:17:08,040 --> 00:17:10,069
such a high performance way and so it
augments our design Engineers it makes
742
00:17:10,069 --> 00:17:10,079
augments our design Engineers it makes
743
00:17:10,079 --> 00:17:11,510
augments our design Engineers it makes
it possible for us to create some of
744
00:17:11,510 --> 00:17:11,520
it possible for us to create some of
745
00:17:11,520 --> 00:17:13,189
it possible for us to create some of
these amazing things at all and of
746
00:17:13,189 --> 00:17:13,199
these amazing things at all and of
747
00:17:13,199 --> 00:17:15,409
these amazing things at all and of
course the productivity of the teams go
748
00:17:15,409 --> 00:17:15,419
course the productivity of the teams go
749
00:17:15,419 --> 00:17:17,750
course the productivity of the teams go
up tremendously well we would like to do
750
00:17:17,750 --> 00:17:17,760
up tremendously well we would like to do
751
00:17:17,760 --> 00:17:19,069
up tremendously well we would like to do
this in just about every single industry
752
00:17:19,069 --> 00:17:19,079
this in just about every single industry
753
00:17:19,079 --> 00:17:20,929
this in just about every single industry
now we just need some powerful machines
754
00:17:20,929 --> 00:17:20,939
now we just need some powerful machines
755
00:17:20,939 --> 00:17:22,730
now we just need some powerful machines
we have powerful machines in the cloud
756
00:17:22,730 --> 00:17:22,740
we have powerful machines in the cloud
757
00:17:22,740 --> 00:17:25,189
we have powerful machines in the cloud
of course dgx cloud has many many uh
758
00:17:25,189 --> 00:17:25,199
of course dgx cloud has many many uh
759
00:17:25,199 --> 00:17:26,870
of course dgx cloud has many many uh
many footprints around the world but
760
00:17:26,870 --> 00:17:26,880
many footprints around the world but
761
00:17:26,880 --> 00:17:28,309
many footprints around the world but
wouldn't it be great if you had a
762
00:17:28,309 --> 00:17:28,319
wouldn't it be great if you had a
763
00:17:28,319 --> 00:17:30,350
wouldn't it be great if you had a
powerful machine under your desk and so
764
00:17:30,350 --> 00:17:30,360
powerful machine under your desk and so
765
00:17:30,360 --> 00:17:32,030
powerful machine under your desk and so
today we're announcing our latest
766
00:17:32,030 --> 00:17:32,040
today we're announcing our latest
767
00:17:32,040 --> 00:17:34,610
today we're announcing our latest
generation Lovelace GPU
768
00:17:34,610 --> 00:17:34,620
generation Lovelace GPU
769
00:17:34,620 --> 00:17:36,710
generation Lovelace GPU
the most powerful GPU we've ever put in
770
00:17:36,710 --> 00:17:36,720
the most powerful GPU we've ever put in
771
00:17:36,720 --> 00:17:38,270
the most powerful GPU we've ever put in
the workstation
772
00:17:38,270 --> 00:17:38,280
the workstation
773
00:17:38,280 --> 00:17:39,650
the workstation
is now
774
00:17:39,650 --> 00:17:39,660
is now
775
00:17:39,660 --> 00:17:41,870
is now
oh gosh darn it
776
00:17:41,870 --> 00:17:41,880
oh gosh darn it
777
00:17:41,880 --> 00:17:43,730
oh gosh darn it
I just put my fingerprints on there did
778
00:17:43,730 --> 00:17:43,740
I just put my fingerprints on there did
779
00:17:43,740 --> 00:17:46,130
I just put my fingerprints on there did
you guys can you guys see that
780
00:17:46,130 --> 00:17:46,140
you guys can you guys see that
781
00:17:46,140 --> 00:17:49,070
you guys can you guys see that
that was what that's not me could you
782
00:17:49,070 --> 00:17:49,080
that was what that's not me could you
783
00:17:49,080 --> 00:17:53,750
that was what that's not me could you
hey can I have this clean in the future
784
00:17:53,750 --> 00:17:53,760
785
00:17:53,760 --> 00:17:55,610
my bad
786
00:17:55,610 --> 00:17:55,620
my bad
787
00:17:55,620 --> 00:17:59,150
my bad
sorry everybody uh my bad
788
00:17:59,150 --> 00:17:59,160
sorry everybody uh my bad
789
00:17:59,160 --> 00:18:01,669
sorry everybody uh my bad
they work so hard it goes into these
790
00:18:01,669 --> 00:18:01,679
they work so hard it goes into these
791
00:18:01,679 --> 00:18:03,950
they work so hard it goes into these
amazing workstations and these amazing
792
00:18:03,950 --> 00:18:03,960
amazing workstations and these amazing
793
00:18:03,960 --> 00:18:05,169
amazing workstations and these amazing
workstations
794
00:18:05,169 --> 00:18:05,179
workstations
795
00:18:05,179 --> 00:18:08,810
workstations
packs up to four of these gpus it packs
796
00:18:08,810 --> 00:18:08,820
packs up to four of these gpus it packs
797
00:18:08,820 --> 00:18:11,210
packs up to four of these gpus it packs
up to four Nvidia RTX 6000 is the most
798
00:18:11,210 --> 00:18:11,220
up to four Nvidia RTX 6000 is the most
799
00:18:11,220 --> 00:18:15,110
up to four Nvidia RTX 6000 is the most
powerful gpus ever created and it run
800
00:18:15,110 --> 00:18:15,120
powerful gpus ever created and it run
801
00:18:15,120 --> 00:18:18,110
powerful gpus ever created and it run
real-time Ray tracing for Omniverse as
802
00:18:18,110 --> 00:18:18,120
real-time Ray tracing for Omniverse as
803
00:18:18,120 --> 00:18:21,770
real-time Ray tracing for Omniverse as
well as train fine tune and inference
804
00:18:21,770 --> 00:18:21,780
well as train fine tune and inference
805
00:18:21,780 --> 00:18:25,029
well as train fine tune and inference
large language models for generative AI
806
00:18:25,029 --> 00:18:25,039
large language models for generative AI
807
00:18:25,039 --> 00:18:27,049
large language models for generative AI
incredibly fast and critically powerful
808
00:18:27,049 --> 00:18:27,059
incredibly fast and critically powerful
809
00:18:27,059 --> 00:18:29,810
incredibly fast and critically powerful
and it produces answers in seconds not
810
00:18:29,810 --> 00:18:29,820
and it produces answers in seconds not
811
00:18:29,820 --> 00:18:32,029
and it produces answers in seconds not
minutes uh for in some of the services
812
00:18:32,029 --> 00:18:32,039
minutes uh for in some of the services
813
00:18:32,039 --> 00:18:34,669
minutes uh for in some of the services
that are out there okay and so another
814
00:18:34,669 --> 00:18:34,679
that are out there okay and so another
815
00:18:34,679 --> 00:18:37,370
that are out there okay and so another
incredible machine are the servers and
816
00:18:37,370 --> 00:18:37,380
incredible machine are the servers and
817
00:18:37,380 --> 00:18:39,590
incredible machine are the servers and
these servers as you know getting gpus
818
00:18:39,590 --> 00:18:39,600
these servers as you know getting gpus
819
00:18:39,600 --> 00:18:41,330
these servers as you know getting gpus
in the cloud these days is no easy feat
820
00:18:41,330 --> 00:18:41,340
in the cloud these days is no easy feat
821
00:18:41,340 --> 00:18:43,610
in the cloud these days is no easy feat
and now you can buy it okay you could
822
00:18:43,610 --> 00:18:43,620
and now you can buy it okay you could
823
00:18:43,620 --> 00:18:45,409
and now you can buy it okay you could
have your your company buy it for you
824
00:18:45,409 --> 00:18:45,419
have your your company buy it for you
825
00:18:45,419 --> 00:18:47,270
have your your company buy it for you
and put it in the data center and
826
00:18:47,270 --> 00:18:47,280
and put it in the data center and
827
00:18:47,280 --> 00:18:49,430
and put it in the data center and
there's a whole bunch of these servers a
828
00:18:49,430 --> 00:18:49,440
there's a whole bunch of these servers a
829
00:18:49,440 --> 00:18:50,690
there's a whole bunch of these servers a
whole bunch of different configurations
830
00:18:50,690 --> 00:18:50,700
whole bunch of different configurations
831
00:18:50,700 --> 00:18:52,130
whole bunch of different configurations
I don't know if you guys could see this
832
00:18:52,130 --> 00:18:52,140
I don't know if you guys could see this
833
00:18:52,140 --> 00:18:54,730
I don't know if you guys could see this
this is a server that has up to eight of
834
00:18:54,730 --> 00:18:54,740
this is a server that has up to eight of
835
00:18:54,740 --> 00:18:57,650
this is a server that has up to eight of
the l40s
836
00:18:57,650 --> 00:18:57,660
the l40s
837
00:18:57,660 --> 00:19:01,490
the l40s
Ada Lovelace gpus and of course these
838
00:19:01,490 --> 00:19:01,500
Ada Lovelace gpus and of course these
839
00:19:01,500 --> 00:19:02,810
Ada Lovelace gpus and of course these
are not going to be used for Frontier
840
00:19:02,810 --> 00:19:02,820
are not going to be used for Frontier
841
00:19:02,820 --> 00:19:05,990
are not going to be used for Frontier
models these are really used for
842
00:19:05,990 --> 00:19:06,000
models these are really used for
843
00:19:06,000 --> 00:19:08,029
models these are really used for
mainstream models today that you can
844
00:19:08,029 --> 00:19:08,039
mainstream models today that you can
845
00:19:08,039 --> 00:19:10,549
mainstream models today that you can
download from hugging face or Nvidia
846
00:19:10,549 --> 00:19:10,559
download from hugging face or Nvidia
847
00:19:10,559 --> 00:19:11,810
download from hugging face or Nvidia
could work with your company to create
848
00:19:11,810 --> 00:19:11,820
could work with your company to create
849
00:19:11,820 --> 00:19:14,810
could work with your company to create
you could use in just about all kinds of
850
00:19:14,810 --> 00:19:14,820
you could use in just about all kinds of
851
00:19:14,820 --> 00:19:17,210
you could use in just about all kinds of
applications around your company and you
852
00:19:17,210 --> 00:19:17,220
applications around your company and you
853
00:19:17,220 --> 00:19:19,970
applications around your company and you
can fine tune it with these gpus the
854
00:19:19,970 --> 00:19:19,980
can fine tune it with these gpus the
855
00:19:19,980 --> 00:19:24,289
can fine tune it with these gpus the
fine tuning of a gpt3 model okay so this
856
00:19:24,289 --> 00:19:24,299
fine tuning of a gpt3 model okay so this
857
00:19:24,299 --> 00:19:27,230
fine tuning of a gpt3 model okay so this
is GPT 340 billion parameters takes
858
00:19:27,230 --> 00:19:27,240
is GPT 340 billion parameters takes
859
00:19:27,240 --> 00:19:29,750
is GPT 340 billion parameters takes
about seven hours for about a billion
860
00:19:29,750 --> 00:19:29,760
about seven hours for about a billion
861
00:19:29,760 --> 00:19:31,669
about seven hours for about a billion
tokens and so 15 hours in a workstation
862
00:19:31,669 --> 00:19:31,679
tokens and so 15 hours in a workstation
863
00:19:31,679 --> 00:19:34,010
tokens and so 15 hours in a workstation
with four gpus of course takes less with
864
00:19:34,010 --> 00:19:34,020
with four gpus of course takes less with
865
00:19:34,020 --> 00:19:37,070
with four gpus of course takes less with
agpus and just in fine tuning this is
866
00:19:37,070 --> 00:19:37,080
agpus and just in fine tuning this is
867
00:19:37,080 --> 00:19:40,010
agpus and just in fine tuning this is
one and a half times faster than our
868
00:19:40,010 --> 00:19:40,020
one and a half times faster than our
869
00:19:40,020 --> 00:19:41,450
one and a half times faster than our
last generation
870
00:19:41,450 --> 00:19:41,460
last generation
871
00:19:41,460 --> 00:19:44,630
last generation
a100 and so l40s is a really terrific
872
00:19:44,630 --> 00:19:44,640
a100 and so l40s is a really terrific
873
00:19:44,640 --> 00:19:48,650
a100 and so l40s is a really terrific
GPU for Enterprise scale fine-tuning of
874
00:19:48,650 --> 00:19:48,660
GPU for Enterprise scale fine-tuning of
875
00:19:48,660 --> 00:19:50,870
GPU for Enterprise scale fine-tuning of
mainstream large language models these
876
00:19:50,870 --> 00:19:50,880
mainstream large language models these
877
00:19:50,880 --> 00:19:52,730
mainstream large language models these
amazing new Enterprise systems that are
878
00:19:52,730 --> 00:19:52,740
amazing new Enterprise systems that are
879
00:19:52,740 --> 00:19:55,310
amazing new Enterprise systems that are
in production today
880
00:19:55,310 --> 00:19:55,320
in production today
881
00:19:55,320 --> 00:19:56,990
in production today
all right let's change gears and talk
882
00:19:56,990 --> 00:19:57,000
all right let's change gears and talk
883
00:19:57,000 --> 00:19:59,090
all right let's change gears and talk
about what's going on at siggraph this
884
00:19:59,090 --> 00:19:59,100
about what's going on at siggraph this
885
00:19:59,100 --> 00:20:00,049
about what's going on at siggraph this
year
886
00:20:00,049 --> 00:20:00,059
year
887
00:20:00,059 --> 00:20:02,510
year
I'm pretty sure all of you have already
888
00:20:02,510 --> 00:20:02,520
I'm pretty sure all of you have already
889
00:20:02,520 --> 00:20:07,850
I'm pretty sure all of you have already
heard about open USD USD openusd is a
890
00:20:07,850 --> 00:20:07,860
heard about open USD USD openusd is a
891
00:20:07,860 --> 00:20:12,770
heard about open USD USD openusd is a
very big deal open USD is a framework
892
00:20:12,770 --> 00:20:12,780
very big deal open USD is a framework
893
00:20:12,780 --> 00:20:16,610
very big deal open USD is a framework
a universal interchange for creating 3D
894
00:20:16,610 --> 00:20:16,620
a universal interchange for creating 3D
895
00:20:16,620 --> 00:20:19,549
a universal interchange for creating 3D
worlds for describing for compositing
896
00:20:19,549 --> 00:20:19,559
worlds for describing for compositing
897
00:20:19,559 --> 00:20:22,250
worlds for describing for compositing
for simulating for collaborating on 3D
898
00:20:22,250 --> 00:20:22,260
for simulating for collaborating on 3D
899
00:20:22,260 --> 00:20:24,830
for simulating for collaborating on 3D
projects
900
00:20:24,830 --> 00:20:24,840
901
00:20:24,840 --> 00:20:26,690
open USD
902
00:20:26,690 --> 00:20:26,700
open USD
903
00:20:26,700 --> 00:20:29,810
open USD
is going to bring together the world
904
00:20:29,810 --> 00:20:29,820
is going to bring together the world
905
00:20:29,820 --> 00:20:34,310
is going to bring together the world
onto one standard 3D interchange
906
00:20:34,310 --> 00:20:34,320
onto one standard 3D interchange
907
00:20:34,320 --> 00:20:36,169
onto one standard 3D interchange
and has the opportunity to do for the
908
00:20:36,169 --> 00:20:36,179
and has the opportunity to do for the
909
00:20:36,179 --> 00:20:40,130
and has the opportunity to do for the
world and for computing what HTML did
910
00:20:40,130 --> 00:20:40,140
world and for computing what HTML did
911
00:20:40,140 --> 00:20:42,350
world and for computing what HTML did
for the 2D web
912
00:20:42,350 --> 00:20:42,360
for the 2D web
913
00:20:42,360 --> 00:20:45,470
for the 2D web
finally an industry standard
914
00:20:45,470 --> 00:20:45,480
finally an industry standard
915
00:20:45,480 --> 00:20:48,230
finally an industry standard
powerful and extensible 3D interchange
916
00:20:48,230 --> 00:20:48,240
powerful and extensible 3D interchange
917
00:20:48,240 --> 00:20:51,710
powerful and extensible 3D interchange
that brings the whole world together a
918
00:20:51,710 --> 00:20:51,720
that brings the whole world together a
919
00:20:51,720 --> 00:20:54,049
that brings the whole world together a
really big deal now let's take a look at
920
00:20:54,049 --> 00:20:54,059
really big deal now let's take a look at
921
00:20:54,059 --> 00:20:55,909
really big deal now let's take a look at
why it is such a Visionary thing that
922
00:20:55,909 --> 00:20:55,919
why it is such a Visionary thing that
923
00:20:55,919 --> 00:20:57,770
why it is such a Visionary thing that
Pixar did I forget exactly when they
924
00:20:57,770 --> 00:20:57,780
Pixar did I forget exactly when they
925
00:20:57,780 --> 00:20:59,150
Pixar did I forget exactly when they
invented it but they open sourced it in
926
00:20:59,150 --> 00:20:59,160
invented it but they open sourced it in
927
00:20:59,160 --> 00:21:02,690
invented it but they open sourced it in
2015 and they've been of course using
928
00:21:02,690 --> 00:21:02,700
2015 and they've been of course using
929
00:21:02,700 --> 00:21:05,210
2015 and they've been of course using
this framework for over a decade
930
00:21:05,210 --> 00:21:05,220
this framework for over a decade
931
00:21:05,220 --> 00:21:08,029
this framework for over a decade
building amazing 3D content well the 3D
932
00:21:08,029 --> 00:21:08,039
building amazing 3D content well the 3D
933
00:21:08,039 --> 00:21:09,710
building amazing 3D content well the 3D
pipeline is incredibly complicated you
934
00:21:09,710 --> 00:21:09,720
pipeline is incredibly complicated you
935
00:21:09,720 --> 00:21:11,210
pipeline is incredibly complicated you
got designers and artists and Engineers
936
00:21:11,210 --> 00:21:11,220
got designers and artists and Engineers
937
00:21:11,220 --> 00:21:13,669
got designers and artists and Engineers
they all specialize in some part of the
938
00:21:13,669 --> 00:21:13,679
they all specialize in some part of the
939
00:21:13,679 --> 00:21:15,590
they all specialize in some part of the
3D workflow it could be modeling and
940
00:21:15,590 --> 00:21:15,600
3D workflow it could be modeling and
941
00:21:15,600 --> 00:21:16,970
3D workflow it could be modeling and
texturing
942
00:21:16,970 --> 00:21:16,980
texturing
943
00:21:16,980 --> 00:21:18,470
texturing
materials
944
00:21:18,470 --> 00:21:18,480
materials
945
00:21:18,480 --> 00:21:20,690
materials
physics simulation
946
00:21:20,690 --> 00:21:20,700
physics simulation
947
00:21:20,700 --> 00:21:22,070
physics simulation
animation
948
00:21:22,070 --> 00:21:22,080
animation
949
00:21:22,080 --> 00:21:25,850
animation
set design scene composition
950
00:21:25,850 --> 00:21:25,860
set design scene composition
951
00:21:25,860 --> 00:21:28,310
set design scene composition
there are so many parts and so many
952
00:21:28,310 --> 00:21:28,320
there are so many parts and so many
953
00:21:28,320 --> 00:21:30,230
there are so many parts and so many
different tools and because the tools
954
00:21:30,230 --> 00:21:30,240
different tools and because the tools
955
00:21:30,240 --> 00:21:32,210
different tools and because the tools
are created by different companies and
956
00:21:32,210 --> 00:21:32,220
are created by different companies and
957
00:21:32,220 --> 00:21:33,950
are created by different companies and
largely incompatible
958
00:21:33,950 --> 00:21:33,960
largely incompatible
959
00:21:33,960 --> 00:21:36,950
largely incompatible
import and exporting data conversion is
960
00:21:36,950 --> 00:21:36,960
import and exporting data conversion is
961
00:21:36,960 --> 00:21:38,750
import and exporting data conversion is
just part of the workflow and because
962
00:21:38,750 --> 00:21:38,760
just part of the workflow and because
963
00:21:38,760 --> 00:21:40,070
just part of the workflow and because
they're incompatible and because there's
964
00:21:40,070 --> 00:21:40,080
they're incompatible and because there's
965
00:21:40,080 --> 00:21:41,870
they're incompatible and because there's
all this import and exporting
966
00:21:41,870 --> 00:21:41,880
all this import and exporting
967
00:21:41,880 --> 00:21:44,090
all this import and exporting
fundamentally the workflow has to be
968
00:21:44,090 --> 00:21:44,100
fundamentally the workflow has to be
969
00:21:44,100 --> 00:21:45,950
fundamentally the workflow has to be
serialized it's impossible to paralyze
970
00:21:45,950 --> 00:21:45,960
serialized it's impossible to paralyze
971
00:21:45,960 --> 00:21:46,610
serialized it's impossible to paralyze
that
972
00:21:46,610 --> 00:21:46,620
that
973
00:21:46,620 --> 00:21:49,250
that
and this is one of the reasons why
974
00:21:49,250 --> 00:21:49,260
and this is one of the reasons why
975
00:21:49,260 --> 00:21:52,610
and this is one of the reasons why
creating these incredible 3D animation
976
00:21:52,610 --> 00:21:52,620
creating these incredible 3D animation
977
00:21:52,620 --> 00:21:54,770
creating these incredible 3D animation
movies are so expensive and take so much
978
00:21:54,770 --> 00:21:54,780
movies are so expensive and take so much
979
00:21:54,780 --> 00:21:56,930
movies are so expensive and take so much
time could you imagine if every single
980
00:21:56,930 --> 00:21:56,940
time could you imagine if every single
981
00:21:56,940 --> 00:22:00,649
time could you imagine if every single
tool was natively compatible with USD
982
00:22:00,649 --> 00:22:00,659
tool was natively compatible with USD
983
00:22:00,659 --> 00:22:02,870
tool was natively compatible with USD
then as a result everybody can work in
984
00:22:02,870 --> 00:22:02,880
then as a result everybody can work in
985
00:22:02,880 --> 00:22:04,070
then as a result everybody can work in
parallel
986
00:22:04,070 --> 00:22:04,080
parallel
987
00:22:04,080 --> 00:22:06,890
parallel
The Interchange and conversion goes away
988
00:22:06,890 --> 00:22:06,900
The Interchange and conversion goes away
989
00:22:06,900 --> 00:22:09,470
The Interchange and conversion goes away
and instead of a serialized model you
990
00:22:09,470 --> 00:22:09,480
and instead of a serialized model you
991
00:22:09,480 --> 00:22:11,930
and instead of a serialized model you
have a paralyzed spoken Hub model
992
00:22:11,930 --> 00:22:11,940
have a paralyzed spoken Hub model
993
00:22:11,940 --> 00:22:15,169
have a paralyzed spoken Hub model
and so this way of doing work of course
994
00:22:15,169 --> 00:22:15,179
and so this way of doing work of course
995
00:22:15,179 --> 00:22:17,270
and so this way of doing work of course
is incredibly appealing and it's one of
996
00:22:17,270 --> 00:22:17,280
is incredibly appealing and it's one of
997
00:22:17,280 --> 00:22:19,190
is incredibly appealing and it's one of
the reasons why the vision of open USD
998
00:22:19,190 --> 00:22:19,200
the reasons why the vision of open USD
999
00:22:19,200 --> 00:22:20,990
the reasons why the vision of open USD
has taken off
1000
00:22:20,990 --> 00:22:21,000
has taken off
1001
00:22:21,000 --> 00:22:23,510
has taken off
it's being adopted in film architecture
1002
00:22:23,510 --> 00:22:23,520
it's being adopted in film architecture
1003
00:22:23,520 --> 00:22:24,850
it's being adopted in film architecture
engineering and construction
1004
00:22:24,850 --> 00:22:24,860
engineering and construction
1005
00:22:24,860 --> 00:22:26,870
engineering and construction
manufacturing and so many different
1006
00:22:26,870 --> 00:22:26,880
manufacturing and so many different
1007
00:22:26,880 --> 00:22:29,750
manufacturing and so many different
fields of Robotics well five years ago
1008
00:22:29,750 --> 00:22:29,760
fields of Robotics well five years ago
1009
00:22:29,760 --> 00:22:32,630
fields of Robotics well five years ago
we started working with Pixar and we
1010
00:22:32,630 --> 00:22:32,640
we started working with Pixar and we
1011
00:22:32,640 --> 00:22:35,870
we started working with Pixar and we
adopted USD as the foundation of
1012
00:22:35,870 --> 00:22:35,880
adopted USD as the foundation of
1013
00:22:35,880 --> 00:22:38,330
adopted USD as the foundation of
Omniverse it's not a tool in itself it's
1014
00:22:38,330 --> 00:22:38,340
Omniverse it's not a tool in itself it's
1015
00:22:38,340 --> 00:22:41,090
Omniverse it's not a tool in itself it's
a connector of tools okay so Omniverse
1016
00:22:41,090 --> 00:22:41,100
a connector of tools okay so Omniverse
1017
00:22:41,100 --> 00:22:43,789
a connector of tools okay so Omniverse
is a connector well let's take a look at
1018
00:22:43,789 --> 00:22:43,799
is a connector well let's take a look at
1019
00:22:43,799 --> 00:22:46,310
is a connector well let's take a look at
held the vision of open USD came came
1020
00:22:46,310 --> 00:22:46,320
held the vision of open USD came came
1021
00:22:46,320 --> 00:22:48,649
held the vision of open USD came came
together and this is just a fantastic
1022
00:22:48,649 --> 00:22:48,659
together and this is just a fantastic
1023
00:22:48,659 --> 00:22:51,350
together and this is just a fantastic
illustration let's starting from the
1024
00:22:51,350 --> 00:22:51,360
illustration let's starting from the
1025
00:22:51,360 --> 00:22:53,810
illustration let's starting from the
left here I think this is a Adobe Stager
1026
00:22:53,810 --> 00:22:53,820
left here I think this is a Adobe Stager
1027
00:22:53,820 --> 00:22:57,409
left here I think this is a Adobe Stager
Houdini this is a modeling system Maya
1028
00:22:57,409 --> 00:22:57,419
Houdini this is a modeling system Maya
1029
00:22:57,419 --> 00:22:59,690
Houdini this is a modeling system Maya
or animation system modeling system this
1030
00:22:59,690 --> 00:22:59,700
or animation system modeling system this
1031
00:22:59,700 --> 00:23:01,010
or animation system modeling system this
is Omniverse
1032
00:23:01,010 --> 00:23:01,020
is Omniverse
1033
00:23:01,020 --> 00:23:02,149
is Omniverse
blender
1034
00:23:02,149 --> 00:23:02,159
blender
1035
00:23:02,159 --> 00:23:06,049
blender
render man Pixar's Minuteman and Unreal
1036
00:23:06,049 --> 00:23:06,059
render man Pixar's Minuteman and Unreal
1037
00:23:06,059 --> 00:23:08,990
render man Pixar's Minuteman and Unreal
Engine from epic a game engine literally
1038
00:23:08,990 --> 00:23:09,000
Engine from epic a game engine literally
1039
00:23:09,000 --> 00:23:11,630
Engine from epic a game engine literally
all open USD
1040
00:23:11,630 --> 00:23:11,640
all open USD
1041
00:23:11,640 --> 00:23:13,730
all open USD
one data set
1042
00:23:13,730 --> 00:23:13,740
one data set
1043
00:23:13,740 --> 00:23:16,669
one data set
ingested into everybody's tools and it
1044
00:23:16,669 --> 00:23:16,679
ingested into everybody's tools and it
1045
00:23:16,679 --> 00:23:18,230
ingested into everybody's tools and it
looks basically the same
1046
00:23:18,230 --> 00:23:18,240
looks basically the same
1047
00:23:18,240 --> 00:23:19,730
looks basically the same
everybody's rendering system is a little
1048
00:23:19,730 --> 00:23:19,740
everybody's rendering system is a little
1049
00:23:19,740 --> 00:23:20,930
everybody's rendering system is a little
different and so the quality of the
1050
00:23:20,930 --> 00:23:20,940
different and so the quality of the
1051
00:23:20,940 --> 00:23:22,370
different and so the quality of the
rendering is is a little different from
1052
00:23:22,370 --> 00:23:22,380
rendering is is a little different from
1053
00:23:22,380 --> 00:23:25,130
rendering is is a little different from
tool to Tool but one data set available
1054
00:23:25,130 --> 00:23:25,140
tool to Tool but one data set available
1055
00:23:25,140 --> 00:23:27,950
tool to Tool but one data set available
and usable by every tool this is the
1056
00:23:27,950 --> 00:23:27,960
and usable by every tool this is the
1057
00:23:27,960 --> 00:23:31,549
and usable by every tool this is the
vision of open USD so incredibly
1058
00:23:31,549 --> 00:23:31,559
vision of open USD so incredibly
1059
00:23:31,559 --> 00:23:32,630
vision of open USD so incredibly
powerful
1060
00:23:32,630 --> 00:23:32,640
powerful
1061
00:23:32,640 --> 00:23:34,490
powerful
and so take a look at this this is the
1062
00:23:34,490 --> 00:23:34,500
and so take a look at this this is the
1063
00:23:34,500 --> 00:23:41,690
and so take a look at this this is the
latest of Omniverse
1064
00:23:41,690 --> 00:23:41,700
1065
00:23:41,700 --> 00:23:48,669
[Music]
1066
00:23:48,669 --> 00:23:48,679
[Music]
1067
00:23:48,679 --> 00:23:54,740
[Music]
thank you
1068
00:23:54,740 --> 00:23:54,750
1069
00:23:54,750 --> 00:24:01,149
[Music]
1070
00:24:01,149 --> 00:24:01,159
1071
00:24:01,159 --> 00:24:08,280
thank you
1072
00:24:08,280 --> 00:24:08,290
1073
00:24:08,290 --> 00:24:31,190
[Music]
1074
00:24:31,190 --> 00:24:31,200
1075
00:24:31,200 --> 00:24:56,330
[Music]
1076
00:24:56,330 --> 00:24:56,340
[Music]
1077
00:24:56,340 --> 00:24:57,990
[Music]
foreign
1078
00:24:57,990 --> 00:24:58,000
foreign
1079
00:24:58,000 --> 00:25:08,149
foreign
[Music]
1080
00:25:08,149 --> 00:25:08,159
1081
00:25:08,159 --> 00:25:11,050
foreign
1082
00:25:11,050 --> 00:25:11,060
foreign
1083
00:25:11,060 --> 00:25:28,140
foreign
[Music]
1084
00:25:28,140 --> 00:25:28,150
1085
00:25:28,150 --> 00:25:36,769
[Music]
1086
00:25:36,769 --> 00:25:36,779
1087
00:25:36,779 --> 00:25:47,470
thank you
1088
00:25:47,470 --> 00:25:47,480
1089
00:25:47,480 --> 00:25:55,010
[Music]
1090
00:25:55,010 --> 00:25:55,020
1091
00:25:55,020 --> 00:26:02,750
doesn't that just make you happy
1092
00:26:02,750 --> 00:26:02,760
1093
00:26:02,760 --> 00:26:05,570
no art all physics now Omniverse as I
1094
00:26:05,570 --> 00:26:05,580
no art all physics now Omniverse as I
1095
00:26:05,580 --> 00:26:06,890
no art all physics now Omniverse as I
mentioned before is not a tool it's a
1096
00:26:06,890 --> 00:26:06,900
mentioned before is not a tool it's a
1097
00:26:06,900 --> 00:26:08,930
mentioned before is not a tool it's a
platform for tools it's not a tool is a
1098
00:26:08,930 --> 00:26:08,940
platform for tools it's not a tool is a
1099
00:26:08,940 --> 00:26:11,149
platform for tools it's not a tool is a
platform for tools okay and so we would
1100
00:26:11,149 --> 00:26:11,159
platform for tools okay and so we would
1101
00:26:11,159 --> 00:26:12,890
platform for tools okay and so we would
like to put Omniverse in as many places
1102
00:26:12,890 --> 00:26:12,900
like to put Omniverse in as many places
1103
00:26:12,900 --> 00:26:14,210
like to put Omniverse in as many places
as possible there's a whole bunch of
1104
00:26:14,210 --> 00:26:14,220
as possible there's a whole bunch of
1105
00:26:14,220 --> 00:26:15,950
as possible there's a whole bunch of
different applications and we showed you
1106
00:26:15,950 --> 00:26:15,960
different applications and we showed you
1107
00:26:15,960 --> 00:26:18,830
different applications and we showed you
earlier how we used AI workbench to
1108
00:26:18,830 --> 00:26:18,840
earlier how we used AI workbench to
1109
00:26:18,840 --> 00:26:21,649
earlier how we used AI workbench to
train this model to fine-tune this model
1110
00:26:21,649 --> 00:26:21,659
train this model to fine-tune this model
1111
00:26:21,659 --> 00:26:25,250
train this model to fine-tune this model
we started with llama 2 and we taught it
1112
00:26:25,250 --> 00:26:25,260
we started with llama 2 and we taught it
1113
00:26:25,260 --> 00:26:28,130
we started with llama 2 and we taught it
we taught it we fine-tuned it for USD
1114
00:26:28,130 --> 00:26:28,140
we taught it we fine-tuned it for USD
1115
00:26:28,140 --> 00:26:31,430
we taught it we fine-tuned it for USD
and so let's take a look at the video
1116
00:26:31,430 --> 00:26:31,440
and so let's take a look at the video
1117
00:26:31,440 --> 00:26:34,130
and so let's take a look at the video
for USD developers building profiling
1118
00:26:34,130 --> 00:26:34,140
for USD developers building profiling
1119
00:26:34,140 --> 00:26:36,529
for USD developers building profiling
and optimizing large 3D scenes can be a
1120
00:26:36,529 --> 00:26:36,539
and optimizing large 3D scenes can be a
1121
00:26:36,539 --> 00:26:38,149
and optimizing large 3D scenes can be a
very complex process
1122
00:26:38,149 --> 00:26:38,159
very complex process
1123
00:26:38,159 --> 00:26:40,789
very complex process
chat USD is an llm that's fine-tuned
1124
00:26:40,789 --> 00:26:40,799
chat USD is an llm that's fine-tuned
1125
00:26:40,799 --> 00:26:43,130
chat USD is an llm that's fine-tuned
with USD functions and python USD code
1126
00:26:43,130 --> 00:26:43,140
with USD functions and python USD code
1127
00:26:43,140 --> 00:26:46,010
with USD functions and python USD code
Snippets using Nvidia AI workbench and
1128
00:26:46,010 --> 00:26:46,020
Snippets using Nvidia AI workbench and
1129
00:26:46,020 --> 00:26:47,630
Snippets using Nvidia AI workbench and
the Nemo framework
1130
00:26:47,630 --> 00:26:47,640
the Nemo framework
1131
00:26:47,640 --> 00:26:50,090
the Nemo framework
this generative AI copilot is easily
1132
00:26:50,090 --> 00:26:50,100
this generative AI copilot is easily
1133
00:26:50,100 --> 00:26:52,269
this generative AI copilot is easily
accessed as an Omniverse Cloud API
1134
00:26:52,269 --> 00:26:52,279
accessed as an Omniverse Cloud API
1135
00:26:52,279 --> 00:26:54,830
accessed as an Omniverse Cloud API
simplifying your USD development tasks
1136
00:26:54,830 --> 00:26:54,840
simplifying your USD development tasks
1137
00:26:54,840 --> 00:26:56,870
simplifying your USD development tasks
directly in Omniverse
1138
00:26:56,870 --> 00:26:56,880
directly in Omniverse
1139
00:26:56,880 --> 00:26:59,750
directly in Omniverse
use chat USD for general knowledge like
1140
00:26:59,750 --> 00:26:59,760
use chat USD for general knowledge like
1141
00:26:59,760 --> 00:27:01,610
use chat USD for general knowledge like
to understand the geometry properties of
1142
00:27:01,610 --> 00:27:01,620
to understand the geometry properties of
1143
00:27:01,620 --> 00:27:04,370
to understand the geometry properties of
your USD schema or complete previously
1144
00:27:04,370 --> 00:27:04,380
your USD schema or complete previously
1145
00:27:04,380 --> 00:27:06,710
your USD schema or complete previously
tedious repetitive tasks
1146
00:27:06,710 --> 00:27:06,720
tedious repetitive tasks
1147
00:27:06,720 --> 00:27:08,810
tedious repetitive tasks
like generating code to find and replace
1148
00:27:08,810 --> 00:27:08,820
like generating code to find and replace
1149
00:27:08,820 --> 00:27:12,909
like generating code to find and replace
materials on specific objects
1150
00:27:12,909 --> 00:27:12,919
1151
00:27:12,919 --> 00:27:16,250
or to instantly expose all variants of a
1152
00:27:16,250 --> 00:27:16,260
or to instantly expose all variants of a
1153
00:27:16,260 --> 00:27:16,860
or to instantly expose all variants of a
USD print
1154
00:27:16,860 --> 00:27:16,870
USD print
1155
00:27:16,870 --> 00:27:21,590
USD print
[Music]
1156
00:27:21,590 --> 00:27:21,600
1157
00:27:21,600 --> 00:27:23,870
chat USD can also help you build complex
1158
00:27:23,870 --> 00:27:23,880
chat USD can also help you build complex
1159
00:27:23,880 --> 00:27:26,510
chat USD can also help you build complex
scenes such as scaling a scene and
1160
00:27:26,510 --> 00:27:26,520
scenes such as scaling a scene and
1161
00:27:26,520 --> 00:27:28,250
scenes such as scaling a scene and
organizing it in a certain way in your
1162
00:27:28,250 --> 00:27:28,260
organizing it in a certain way in your
1163
00:27:28,260 --> 00:27:31,250
organizing it in a certain way in your
USD stage
1164
00:27:31,250 --> 00:27:31,260
1165
00:27:31,260 --> 00:27:33,890
build bigger more complex Virtual Worlds
1166
00:27:33,890 --> 00:27:33,900
build bigger more complex Virtual Worlds
1167
00:27:33,900 --> 00:27:35,690
build bigger more complex Virtual Worlds
faster than ever with chat USB
1168
00:27:35,690 --> 00:27:35,700
faster than ever with chat USB
1169
00:27:35,700 --> 00:27:39,230
faster than ever with chat USB
generative AI for USD workflows
1170
00:27:39,230 --> 00:27:39,240
generative AI for USD workflows
1171
00:27:39,240 --> 00:27:42,590
generative AI for USD workflows
chat USD now everybody can speak USD and
1172
00:27:42,590 --> 00:27:42,600
chat USD now everybody can speak USD and
1173
00:27:42,600 --> 00:27:44,990
chat USD now everybody can speak USD and
chat USB USD could be a USD teacher it
1174
00:27:44,990 --> 00:27:45,000
chat USB USD could be a USD teacher it
1175
00:27:45,000 --> 00:27:46,909
chat USB USD could be a USD teacher it
could be a USD copilot and help you
1176
00:27:46,909 --> 00:27:46,919
could be a USD copilot and help you
1177
00:27:46,919 --> 00:27:49,430
could be a USD copilot and help you
create your virtual world okay enhance
1178
00:27:49,430 --> 00:27:49,440
create your virtual world okay enhance
1179
00:27:49,440 --> 00:27:51,529
create your virtual world okay enhance
your productivity incredibly we're super
1180
00:27:51,529 --> 00:27:51,539
your productivity incredibly we're super
1181
00:27:51,539 --> 00:27:52,970
your productivity incredibly we're super
excited about the work that we're doing
1182
00:27:52,970 --> 00:27:52,980
excited about the work that we're doing
1183
00:27:52,980 --> 00:27:55,149
excited about the work that we're doing
here we're so happy that we chose
1184
00:27:55,149 --> 00:27:55,159
here we're so happy that we chose
1185
00:27:55,159 --> 00:27:58,310
here we're so happy that we chose
openusd as the foundation of Omniverse
1186
00:27:58,310 --> 00:27:58,320
openusd as the foundation of Omniverse
1187
00:27:58,320 --> 00:28:00,049
openusd as the foundation of Omniverse
and all of the work that we've done to
1188
00:28:00,049 --> 00:28:00,059
and all of the work that we've done to
1189
00:28:00,059 --> 00:28:01,990
and all of the work that we've done to
extend it into real time into
1190
00:28:01,990 --> 00:28:02,000
extend it into real time into
1191
00:28:02,000 --> 00:28:04,730
extend it into real time into
physics-based applications and this is
1192
00:28:04,730 --> 00:28:04,740
physics-based applications and this is
1193
00:28:04,740 --> 00:28:06,710
physics-based applications and this is
the beginning of a journey that we will
1194
00:28:06,710 --> 00:28:06,720
the beginning of a journey that we will
1195
00:28:06,720 --> 00:28:09,110
the beginning of a journey that we will
finally be able to digitalize
1196
00:28:09,110 --> 00:28:09,120
finally be able to digitalize
1197
00:28:09,120 --> 00:28:11,690
finally be able to digitalize
to bring software-driven artificial
1198
00:28:11,690 --> 00:28:11,700
to bring software-driven artificial
1199
00:28:11,700 --> 00:28:14,570
to bring software-driven artificial
intelligence powered workflows into the
1200
00:28:14,570 --> 00:28:14,580
intelligence powered workflows into the
1201
00:28:14,580 --> 00:28:17,029
intelligence powered workflows into the
world's heavy Industries the 50 trillion
1202
00:28:17,029 --> 00:28:17,039
world's heavy Industries the 50 trillion
1203
00:28:17,039 --> 00:28:18,649
world's heavy Industries the 50 trillion
dollars worth of industries that are
1204
00:28:18,649 --> 00:28:18,659
dollars worth of industries that are
1205
00:28:18,659 --> 00:28:20,510
dollars worth of industries that are
wasting enormous amounts of energy and
1206
00:28:20,510 --> 00:28:20,520
wasting enormous amounts of energy and
1207
00:28:20,520 --> 00:28:23,090
wasting enormous amounts of energy and
money and time all the time because it
1208
00:28:23,090 --> 00:28:23,100
money and time all the time because it
1209
00:28:23,100 --> 00:28:25,070
money and time all the time because it
was simply based built on technology
1210
00:28:25,070 --> 00:28:25,080
was simply based built on technology
1211
00:28:25,080 --> 00:28:26,930
was simply based built on technology
that wasn't available at the time and so
1212
00:28:26,930 --> 00:28:26,940
that wasn't available at the time and so
1213
00:28:26,940 --> 00:28:30,289
that wasn't available at the time and so
Omniverse for industrial digitalization
1214
00:28:30,289 --> 00:28:30,299
Omniverse for industrial digitalization
1215
00:28:30,299 --> 00:28:32,750
Omniverse for industrial digitalization
well all of this momentum that we've
1216
00:28:32,750 --> 00:28:32,760
well all of this momentum that we've
1217
00:28:32,760 --> 00:28:34,850
well all of this momentum that we've
already seen with openusd is about to
1218
00:28:34,850 --> 00:28:34,860
already seen with openusd is about to
1219
00:28:34,860 --> 00:28:37,490
already seen with openusd is about to
get turbocharged
1220
00:28:37,490 --> 00:28:37,500
get turbocharged
1221
00:28:37,500 --> 00:28:41,750
get turbocharged
Alliance for open USD was announced with
1222
00:28:41,750 --> 00:28:41,760
Alliance for open USD was announced with
1223
00:28:41,760 --> 00:28:45,950
Alliance for open USD was announced with
Pixar Apple Adobe Autodesk Nvidia as the
1224
00:28:45,950 --> 00:28:45,960
Pixar Apple Adobe Autodesk Nvidia as the
1225
00:28:45,960 --> 00:28:47,870
Pixar Apple Adobe Autodesk Nvidia as the
founding members the alliance's mission
1226
00:28:47,870 --> 00:28:47,880
founding members the alliance's mission
1227
00:28:47,880 --> 00:28:49,909
founding members the alliance's mission
is to Foster development and
1228
00:28:49,909 --> 00:28:49,919
is to Foster development and
1229
00:28:49,919 --> 00:28:52,190
is to Foster development and
standardization of open USD and
1230
00:28:52,190 --> 00:28:52,200
standardization of open USD and
1231
00:28:52,200 --> 00:28:53,990
standardization of open USD and
accelerate this adoption so whatever
1232
00:28:53,990 --> 00:28:54,000
accelerate this adoption so whatever
1233
00:28:54,000 --> 00:28:55,430
accelerate this adoption so whatever
momentum we've already enjoyed the
1234
00:28:55,430 --> 00:28:55,440
momentum we've already enjoyed the
1235
00:28:55,440 --> 00:28:56,930
momentum we've already enjoyed the
vision that we've already enjoyed it's
1236
00:28:56,930 --> 00:28:56,940
vision that we've already enjoyed it's
1237
00:28:56,940 --> 00:28:59,029
vision that we've already enjoyed it's
about get kicked into Turbo Charge
1238
00:28:59,029 --> 00:28:59,039
about get kicked into Turbo Charge
1239
00:28:59,039 --> 00:29:01,070
about get kicked into Turbo Charge
well I want to thank all of you for
1240
00:29:01,070 --> 00:29:01,080
well I want to thank all of you for
1241
00:29:01,080 --> 00:29:03,190
well I want to thank all of you for
coming today and remember remember
1242
00:29:03,190 --> 00:29:03,200
coming today and remember remember
1243
00:29:03,200 --> 00:29:05,750
coming today and remember remember
accelerated Computing and generative AI
1244
00:29:05,750 --> 00:29:05,760
accelerated Computing and generative AI
1245
00:29:05,760 --> 00:29:08,250
accelerated Computing and generative AI
the more you buy the more you save
1246
00:29:08,250 --> 00:29:08,260
the more you buy the more you save
1247
00:29:08,260 --> 00:29:09,480
the more you buy the more you save
[Applause]
1248
00:29:09,480 --> 00:29:09,490
[Applause]
1249
00:29:09,490 --> 00:29:15,519
[Applause]
[Music]
107017
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.