ABOUT THE SPEAKER
George Tulevski - Materials scientist
IBM's George Tulevski wants to use carbon nanotubes to revolutionize microchip design.

Why you should listen

George Tulevski researches nanomaterials and develops new methods to utilize these materials in technologically relevant applications. He is currently a Research Staff Member at IBM Thomas J. Watson Research Center.

Dr. Tulevski's current work includes advancements in carbon nanotubes that can be used in next-generation computer processors, flexible electronics and sensors. He has co-authored more than 50 peer-reviewed scientific publications and over 40 patents. In addition to his research, Tulevski teaches tomorrow’s nanoscientists at Columbia University, where he received his Ph.D.

More profile about the speaker
George Tulevski | Speaker | TED.com
TED@IBM

George Tulevski: The next step in nanotechnology

Filmed:
1,627,952 views

Nearly every other year the transistors that power silicon computer chip shrink in size by half and double in performance, enabling our devices to become more mobile and accessible. But what happens when these components can't get any smaller? George Tulevski researches the unseen and untapped world of nanomaterials. His current work: developing chemical processes to compel billions of carbon nanotubes to assemble themselves into the patterns needed to build circuits, much the same way natural organisms build intricate, diverse and elegant structures. Could they hold the secret to the next generation of computing?
- Materials scientist
IBM's George Tulevski wants to use carbon nanotubes to revolutionize microchip design. Full bio

Double-click the English transcript below to play the video.

00:12
Let's imagine a sculptor
building a statue,
0
760
2975
00:15
just chipping away with his chisel.
1
3760
2376
00:18
Michelangelo had this elegant way
of describing it when he said,
2
6160
3056
00:21
"Every block of stone
has a statue inside of it,
3
9240
2896
00:24
and it's the task
of the sculptor to discover it."
4
12160
3096
00:27
But what if he worked
in the opposite direction?
5
15280
2416
00:29
Not from a solid block of stone,
6
17720
1856
00:31
but from a pile of dust,
7
19600
1696
00:33
somehow gluing millions of these particles
together to form a statue.
8
21320
4376
00:37
I know that's an absurd notion.
9
25720
1496
00:39
It's probably impossible.
10
27240
1736
00:41
The only way you get
a statue from a pile of dust
11
29000
2936
00:43
is if the statue built itself --
12
31960
2336
00:46
if somehow we could compel millions
of these particles to come together
13
34320
4136
00:50
to form the statue.
14
38480
1280
00:52
Now, as odd as that sounds,
15
40320
1536
00:53
that is almost exactly the problem
I work on in my lab.
16
41880
4136
00:58
I don't build with stone,
17
46040
1416
00:59
I build with nanomaterials.
18
47480
1376
01:00
They're these just impossibly small,
fascinating little objects.
19
48880
4376
01:05
They're so small that if this controller
was a nanoparticle,
20
53280
3296
01:08
a human hair would be the size
of this entire room.
21
56600
2936
01:11
And they're at the heart of a field
we call nanotechnology,
22
59560
2816
01:14
which I'm sure we've all heard about,
23
62400
1815
01:16
and we've all heard
how it is going to change everything.
24
64239
3457
01:19
When I was a graduate student,
25
67720
1496
01:21
it was one of the most exciting times
to be working in nanotechnology.
26
69240
3336
01:24
There were scientific breakthroughs
happening all the time.
27
72600
3216
01:27
The conferences were buzzing,
28
75840
1416
01:29
there was tons of money
pouring in from funding agencies.
29
77280
2680
01:32
And the reason is
30
80760
1256
01:34
when objects get really small,
31
82040
1816
01:35
they're governed by a different set
of physics that govern ordinary objects,
32
83880
3616
01:39
like the ones we interact with.
33
87520
1496
01:41
We call this physics quantum mechanics.
34
89040
1936
01:43
And what it tells you is
that you can precisely tune their behavior
35
91000
3176
01:46
just by making seemingly
small changes to them,
36
94200
2296
01:48
like adding or removing
a handful of atoms,
37
96520
2616
01:51
or twisting the material.
38
99160
1696
01:52
It's like this ultimate toolkit.
39
100880
1776
01:54
You really felt empowered;
you felt like you could make anything.
40
102680
3096
01:57
And we were doing it --
41
105800
1256
01:59
and by we I mean my whole
generation of graduate students.
42
107080
3096
02:02
We were trying to make blazing fast
computers using nanomaterials.
43
110200
3496
02:05
We were constructing quantum dots
44
113720
1616
02:07
that could one day go in your body
and find and fight disease.
45
115360
3376
02:10
There were even groups
trying to make an elevator to space
46
118760
3016
02:13
using carbon nanotubes.
47
121800
1240
02:15
You can look that up, that's true.
48
123600
2400
02:18
Anyways, we thought it was going to affect
49
126776
2000
02:20
all parts of science and technology,
from computing to medicine.
50
128800
3096
02:23
And I have to admit,
51
131920
1256
02:25
I drank all of the Kool-Aid.
52
133200
1976
02:27
I mean, every last drop.
53
135200
2640
02:30
But that was 15 years ago,
54
138520
1800
02:33
and --
55
141160
1216
02:34
fantastic science was done,
really important work.
56
142400
2376
02:36
We've learned a lot.
57
144800
1256
02:38
We were never able to translate
that science into new technologies --
58
146080
4376
02:42
into technologies
that could actually impact people.
59
150480
2776
02:45
And the reason is, these nanomaterials --
60
153280
2216
02:47
they're like a double-edged sword.
61
155520
1656
02:49
The same thing that makes
them so interesting --
62
157200
2256
02:51
their small size --
63
159480
1296
02:52
also makes them impossible to work with.
64
160800
2296
02:55
It's literally like trying to build
a statue out of a pile of dust.
65
163120
3736
02:58
And we just don't have the tools
that are small enough to work with them.
66
166880
3696
03:02
But even if we did,
it wouldn't really matter,
67
170600
2296
03:04
because we couldn't one by one
place millions of particles together
68
172920
3896
03:08
to build a technology.
69
176840
1360
03:10
So because of that,
70
178760
1216
03:12
all of the promise
and all of the excitement
71
180000
2096
03:14
has remained just that:
promise and excitement.
72
182120
2776
03:16
We don't have any
disease-fighting nanobots,
73
184920
2376
03:19
there's no elevators to space,
74
187320
2056
03:21
and the thing that I'm most interested in,
no new types of computing.
75
189400
3696
03:25
Now that last one,
that's a really important one.
76
193120
2536
03:27
We just have come to expect
77
195680
1336
03:29
the pace of computing advancements
to go on indefinitely.
78
197040
3776
03:32
We've built entire economies on this idea.
79
200840
2576
03:35
And this pace exists
80
203440
1736
03:37
because of our ability
to pack more and more devices
81
205200
2456
03:39
onto a computer chip.
82
207680
1696
03:41
And as those devices get smaller,
83
209400
1816
03:43
they get faster, they consume less power
84
211240
2256
03:45
and they get cheaper.
85
213520
1416
03:46
And it's this convergence
that gives us this incredible pace.
86
214960
4576
03:51
As an example:
87
219560
1216
03:52
if I took the room-sized computer
that sent three men to the moon and back
88
220800
5416
03:58
and somehow compressed it --
89
226240
2136
04:00
compressed the world's
greatest computer of its day,
90
228400
3256
04:03
so it was the same size
as your smartphone --
91
231680
2576
04:06
your actual smartphone,
92
234280
1256
04:07
that thing you spent 300 bucks on
and just toss out every two years,
93
235560
3376
04:10
would blow this thing away.
94
238960
2456
04:13
You would not be impressed.
95
241440
1336
04:14
It couldn't do anything
that your smartphone does.
96
242800
2496
04:17
It would be slow,
97
245320
1496
04:18
you couldn't put any of your stuff on it,
98
246840
2176
04:21
you could possibly
get through the first two minutes
99
249040
2456
04:23
of a "Walking Dead" episode
if you're lucky --
100
251520
2176
04:25
(Laughter)
101
253720
1016
04:26
The point is the progress --
it's not gradual.
102
254760
2175
04:28
The progress is relentless.
103
256959
1697
04:30
It's exponential.
104
258680
1255
04:31
It compounds on itself year after year,
105
259959
2337
04:34
to the point where
if you compare a technology
106
262320
2176
04:36
from one generation to the next,
107
264520
1696
04:38
they're almost unrecognizable.
108
266240
1976
04:40
And we owe it to ourselves
to keep this progress going.
109
268240
2616
04:42
We want to say the same thing
10, 20, 30 years from now:
110
270880
3616
04:46
look what we've done
over the last 30 years.
111
274520
2080
04:49
Yet we know this progress
may not last forever.
112
277200
2736
04:51
In fact, the party's kind of winding down.
113
279960
2056
04:54
It's like "last call for alcohol," right?
114
282040
2336
04:56
If you look under the covers,
115
284400
1656
04:58
by many metrics
like speed and performance,
116
286080
2576
05:00
the progress has already slowed to a halt.
117
288680
2520
05:03
So if we want to keep this party going,
118
291760
2136
05:05
we have to do what we've
always been able to do,
119
293920
2256
05:08
and that is to innovate.
120
296200
1456
05:09
So our group's role
and our group's mission
121
297680
2576
05:12
is to innovate
by employing carbon nanotubes,
122
300280
2416
05:14
because we think that they can
provide a path to continue this pace.
123
302720
4056
05:18
They are just like they sound.
124
306800
1456
05:20
They're tiny, hollow tubes
of carbon atoms,
125
308280
2456
05:22
and their nanoscale size,
that small size,
126
310760
2936
05:25
gives rise to these
just outstanding electronic properties.
127
313720
3656
05:29
And the science tells us
if we could employ them in computing,
128
317400
3656
05:33
we could see up to a ten times
improvement in performance.
129
321080
2736
05:35
It's like skipping through several
technology generations in just one step.
130
323840
4400
05:40
So there we have it.
131
328840
1256
05:42
We have this really important problem
132
330120
2016
05:44
and we have what is basically
the ideal solution.
133
332160
2376
05:46
The science is screaming at us,
134
334560
1656
05:48
"This is what you should be doing
to solve your problem."
135
336240
2840
05:53
So, all right, let's get started,
136
341480
1616
05:55
let's do this.
137
343120
1256
05:56
But you just run right back
into that double-edged sword.
138
344400
2696
05:59
This "ideal solution" contains a material
that's impossible to work with.
139
347120
3576
06:02
I'd have to arrange billions of them
just to make one single computer chip.
140
350720
4296
06:07
It's that same conundrum,
it's like this undying problem.
141
355040
3600
06:11
At this point, we said, "Let's just stop.
142
359360
1976
06:13
Let's not go down that same road.
143
361360
1936
06:15
Let's just figure out what's missing.
144
363320
2536
06:17
What are we not dealing with?
145
365880
1416
06:19
What are we not doing
that needs to be done?"
146
367320
2136
06:21
It's like in "The Godfather," right?
147
369480
1776
06:23
When Fredo betrays his brother Michael,
148
371280
2296
06:25
we all know what needs to be done.
149
373600
1656
06:27
Fredo's got to go.
150
375280
1336
06:28
(Laughter)
151
376640
1016
06:29
But Michael -- he puts it off.
152
377680
1936
06:31
Fine, I get it.
153
379640
1216
06:32
Their mother's still alive,
it would make her upset.
154
380880
2456
06:35
We just said,
155
383360
1416
06:36
"What's the Fredo in our problem?"
156
384800
2296
06:39
What are we not dealing with?
157
387120
1416
06:40
What are we not doing,
158
388560
1536
06:42
but needs to be done
to make this a success?"
159
390120
2520
06:45
And the answer is
that the statue has to build itself.
160
393200
4056
06:49
We have to find a way, somehow,
161
397280
1936
06:51
to compel, to convince
billions of these particles
162
399240
4096
06:55
to assemble themselves
into the technology.
163
403360
2976
06:58
We can't do it for them.
They have to do it for themselves.
164
406360
3176
07:01
And it's the hard way,
and this is not trivial,
165
409560
3136
07:04
but in this case, it's the only way.
166
412720
2856
07:07
Now, as it turns out,
this is not that alien of a problem.
167
415600
3896
07:11
We just don't build anything this way.
168
419520
1856
07:13
People don't build anything this way.
169
421400
2016
07:15
But if you look around --
and there's examples everywhere --
170
423440
3176
07:18
Mother Nature builds everything this way.
171
426640
2896
07:21
Everything is built from the bottom up.
172
429560
2656
07:24
You can go to the beach,
173
432240
1256
07:25
you'll find these simple organisms
that use proteins --
174
433520
3176
07:28
basically molecules --
175
436720
1256
07:30
to template what is essentially sand,
176
438000
2056
07:32
just plucking it from the sea
177
440080
1416
07:33
and building these extraordinary
architectures with extreme diversity.
178
441520
3376
07:36
And nature's not crude like us,
just hacking away.
179
444920
2616
07:39
She's elegant and smart,
180
447560
1696
07:41
building with what's available,
molecule by molecule,
181
449280
2736
07:44
making structures with a complexity
182
452040
2136
07:46
and a diversity
that we can't even approach.
183
454200
2280
07:49
And she's already at the nano.
184
457320
1696
07:51
She's been there
for hundreds of millions of years.
185
459040
2496
07:53
We're the ones that are late to the party.
186
461560
2040
07:56
So we decided that we're going
to use the same tool that nature uses,
187
464120
4176
08:00
and that's chemistry.
188
468320
1240
08:02
Chemistry is the missing tool.
189
470000
1456
08:03
And chemistry works in this case
190
471480
2056
08:05
because these nanoscale objects
are about the same size as molecules,
191
473560
3816
08:09
so we can use them
to steer these objects around,
192
477400
2616
08:12
much like a tool.
193
480040
1200
08:13
That's exactly what we've done in our lab.
194
481760
2056
08:15
We've developed chemistry
that goes into the pile of dust,
195
483840
3176
08:19
into the pile of nanoparticles,
196
487040
1496
08:20
and pulls out exactly the ones we need.
197
488560
2376
08:22
Then we can use chemistry to arrange
literally billions of these particles
198
490960
3616
08:26
into the pattern
we need to build circuits.
199
494600
2856
08:29
And because we can do that,
200
497480
1336
08:30
we can build circuits
that are many times faster
201
498840
2376
08:33
than what anyone's been able
to make using nanomaterials before.
202
501240
3216
08:36
Chemistry's the missing tool,
203
504480
1416
08:37
and every day our tool gets sharper
and gets more precise.
204
505920
3616
08:41
And eventually --
205
509560
1216
08:42
and we hope this is
within a handful of years --
206
510800
2616
08:45
we can deliver on one
of those original promises.
207
513440
3376
08:48
Now, computing is just one example.
208
516840
2055
08:50
It's the one that I'm interested in,
that my group is really invested in,
209
518919
3537
08:54
but there are others
in renewable energy, in medicine,
210
522480
3816
08:58
in structural materials,
211
526320
1736
09:00
where the science is going to tell you
to move towards the nano.
212
528080
3056
09:03
That's where the biggest benefit is.
213
531160
2120
09:05
But if we're going to do that,
214
533920
1456
09:07
the scientists of today and tomorrow
are going to need new tools --
215
535400
3176
09:10
tools just like the ones I described.
216
538600
2136
09:12
And they will need chemistry.
That's the point.
217
540760
3856
09:16
The beauty of science is that
once you develop these new tools,
218
544640
3656
09:20
they're out there.
219
548320
1256
09:21
They're out there forever,
220
549600
1256
09:22
and anyone anywhere
can pick them up and use them,
221
550880
2856
09:25
and help to deliver
on the promise of nanotechnology.
222
553760
2960
09:29
Thank you so much for your time.
I appreciate it.
223
557320
2696
09:32
(Applause)
224
560040
2440

▲Back to top

ABOUT THE SPEAKER
George Tulevski - Materials scientist
IBM's George Tulevski wants to use carbon nanotubes to revolutionize microchip design.

Why you should listen

George Tulevski researches nanomaterials and develops new methods to utilize these materials in technologically relevant applications. He is currently a Research Staff Member at IBM Thomas J. Watson Research Center.

Dr. Tulevski's current work includes advancements in carbon nanotubes that can be used in next-generation computer processors, flexible electronics and sensors. He has co-authored more than 50 peer-reviewed scientific publications and over 40 patents. In addition to his research, Tulevski teaches tomorrow’s nanoscientists at Columbia University, where he received his Ph.D.

More profile about the speaker
George Tulevski | Speaker | TED.com