رونوشت 🔗
1
00:00:00,000 --> 00:00:03,870
[MUSIC]
2
00:00:03,870 --> 00:00:07,240
This is Engenuity,
a Boilermaker Engineering podcast.
3
00:00:07,240 --> 00:00:11,072
This episode will discuss building
a semiconductor ecosystem.
4
00:00:11,072 --> 00:00:13,483
[MUSIC]
5
00:00:13,483 --> 00:00:17,953
>> Welcome to the Engenuity podcast,
featuring some of the latest inspiring and
6
00:00:17,953 --> 00:00:21,094
innovative stories from
Purdue Engineering.
7
00:00:21,094 --> 00:00:25,214
I'm Arvind Raman, Boilermaker alumnus and
Dean of Engineering.
8
00:00:25,214 --> 00:00:28,349
Today's episode deals
with semiconductors and
9
00:00:28,349 --> 00:00:32,984
chips which are at the heart of
everything humans do these days.
10
00:00:32,984 --> 00:00:36,492
From the first working
transistor at the Bell labs in
11
00:00:36,492 --> 00:00:41,704
1947 to the billions of transistors
in the chip of an iPhone today.
12
00:00:41,704 --> 00:00:46,294
Modern chips are some of the most
fiendishly complex engineered
13
00:00:46,294 --> 00:00:47,828
devices ever built.
14
00:00:47,828 --> 00:00:50,864
And they're getting even more complex.
15
00:00:50,864 --> 00:00:55,730
On one hand, semiconductor engineers have
come up with ingenious ways to roughly
16
00:00:55,730 --> 00:01:00,240
keep pace with doubling the number of
transistors on a chip every two years,
17
00:01:00,240 --> 00:01:02,324
also known as Moore's law.
18
00:01:02,324 --> 00:01:07,381
On the other hand, the race is on for
chips with mind boggling power that
19
00:01:07,381 --> 00:01:12,115
are needed to fuel the dawn of
the age of artificial intelligence.
20
00:01:12,115 --> 00:01:13,807
And to top that all off,
21
00:01:13,807 --> 00:01:19,222
precisely because chips are of such
critical national security interest,
22
00:01:19,222 --> 00:01:24,565
the Chips Act aims to onshore or
near shore semiconductor manufacturing.
23
00:01:24,565 --> 00:01:29,671
This is helping attract semiconductor
fab plants to West Lafayette,
24
00:01:29,671 --> 00:01:31,321
including MediaTek.
25
00:01:31,321 --> 00:01:36,115
And the most recently planned several
billion dollar investment in an advanced
26
00:01:36,115 --> 00:01:38,405
chip packaging facility by SK Hynix,
27
00:01:38,405 --> 00:01:42,070
which dominates the high bandwidth memory,
or HBM market.
28
00:01:42,070 --> 00:01:48,354
And whose chips get bundled with Nvidia's
most advanced graphic processor units.
29
00:01:48,354 --> 00:01:54,156
Purdue is America's semiconductor
university, and to help unpack all this,
30
00:01:54,156 --> 00:01:59,954
my guest today is Mark Lundstrom,
Purdue's chief semiconductor officer.
31
00:01:59,954 --> 00:02:01,656
Doctor Lundstrom is the Don and
32
00:02:01,656 --> 00:02:06,251
Carol Scifres Distinguished Professor
of Electrical and Computer Engineering.
33
00:02:06,251 --> 00:02:09,566
And a member of the US National Academy
of Engineering, and
34
00:02:09,566 --> 00:02:13,354
the American Association for
the Advancement of Science.
35
00:02:13,354 --> 00:02:17,977
He previously worked at Hewlett Packard
on semiconductor process development and
36
00:02:17,977 --> 00:02:19,514
manufacturing.
37
00:02:19,514 --> 00:02:22,674
Welcome, Doctor Lundstrom, to Engenuity.
38
00:02:22,674 --> 00:02:24,763
>> Thank you, Arvind,
it's always a pleasure.
39
00:02:24,763 --> 00:02:29,031
I'm always happy to talk about
semiconductors with another Boilermaker
40
00:02:29,031 --> 00:02:29,661
alumnus.
41
00:02:29,661 --> 00:02:31,189
>> [LAUGH] That's great.
42
00:02:31,189 --> 00:02:36,101
Well, Doctor Lundstrom,
if you could just get started
43
00:02:36,101 --> 00:02:39,980
telling us a little bit about Moore's law.
44
00:02:39,980 --> 00:02:44,421
Which states that the number
of transistors on each chip
45
00:02:44,421 --> 00:02:48,869
doubles roughly every two years,
is that still valid?
46
00:02:48,869 --> 00:02:50,919
>> Well, that's an interesting question.
47
00:02:50,919 --> 00:02:52,568
>> Yeah.
>> And a lot of people are talking
48
00:02:52,568 --> 00:02:55,359
about it,
there's a little bit of a debate going on.
49
00:02:58,457 --> 00:03:01,239
We're still putting more and
more transistors on a chip.
50
00:03:01,239 --> 00:03:05,743
It's getting harder and harder, it's
taking a little longer than two years.
51
00:03:05,743 --> 00:03:09,716
But the people that are closest to this
can see a way forward, at least for
52
00:03:09,716 --> 00:03:13,579
another 10 years, to continue
putting more transistors on a chip.
53
00:03:13,579 --> 00:03:17,312
Maybe it'll take three years to double,
maybe it'll take four.
54
00:03:17,312 --> 00:03:20,697
And engineers tend to be pretty
conservative when they make
55
00:03:20,697 --> 00:03:22,072
their projections.
56
00:03:22,072 --> 00:03:26,917
If they say we can do this for
another 10 years, we'll probably do it for
57
00:03:26,917 --> 00:03:28,337
another 20 years.
58
00:03:28,337 --> 00:03:33,963
>> That's great, and as you see
[COUGH] advances being made currently,
59
00:03:33,963 --> 00:03:40,420
what's the role of advanced packaging
to help meet Moore's law these days?
60
00:03:40,420 --> 00:03:45,250
>> So people are looking at advanced
packaging really as the way to continue to
61
00:03:45,250 --> 00:03:47,627
drive progress in electronics, and
62
00:03:47,627 --> 00:03:53,124
feed this voracious appetite that AI
has for more computing and more memory.
63
00:03:53,124 --> 00:03:54,569
Moore's law is slowing,
64
00:03:54,569 --> 00:03:59,164
we need to look for other ways to advance
the performance of electronic systems.
65
00:03:59,164 --> 00:04:03,889
So advanced packaging is all about not
putting more transistors on a chip, but
66
00:04:03,889 --> 00:04:05,902
putting more chips in a package.
67
00:04:05,902 --> 00:04:10,638
And putting them in such intimate
contact that it is almost as though
68
00:04:10,638 --> 00:04:13,584
they function as one very large chip.
69
00:04:13,584 --> 00:04:15,115
>> So that's interesting,
70
Load full transcript
00:04:15,115 --> 00:04:19,510
we're able to meet the demands without
necessarily beating the physics.
71
00:04:19,510 --> 00:04:20,228
>> That's right.
72
00:04:20,228 --> 00:04:23,038
And we'll probably start
counting the number of
73
00:04:23,038 --> 00:04:27,396
transistors in a package instead of
the number of transistors on a chip.
74
00:04:27,396 --> 00:04:31,474
And we'll have trillions of
transistors in a package.
75
00:04:31,474 --> 00:04:34,890
>> Trillions of transistors,
mind boggling indeed.
76
00:04:34,890 --> 00:04:39,307
Doctor Lundstrom,
why is it that high end semiconductor
77
00:04:39,307 --> 00:04:42,903
manufacturing mostly happens overseas now?
78
00:04:42,903 --> 00:04:46,066
>> Well, it's not only high-end,
it's also low-end.
79
00:04:46,066 --> 00:04:52,883
So something like 90% of our advanced
semiconductor chip manufacturing
80
00:04:52,883 --> 00:04:59,306
has been outsourced overseas to
sophisticated companies like TSMC.
81
00:04:59,306 --> 00:05:03,533
It just made good business sense,
it costs about 30% less to
82
00:05:03,533 --> 00:05:08,004
build one of these factories
overseas than it does in the US.
83
00:05:08,004 --> 00:05:10,639
And over the decades that
we have been doing this,
84
00:05:10,639 --> 00:05:13,564
they have really honed
their manufacturing skills.
85
00:05:13,564 --> 00:05:15,740
So they're extremely good at this.
86
00:05:15,740 --> 00:05:20,404
Now, we've also outsourced 98% of
the low-end, the packaging of the chips,
87
00:05:20,404 --> 00:05:24,347
not the advanced packaging we've
been talking about for the future.
88
00:05:24,347 --> 00:05:29,120
But the low-end, because that's a very
labor intensive, low margin business,
89
00:05:29,120 --> 00:05:32,576
just made good business sense
to outsource that as well.
90
00:05:32,576 --> 00:05:35,062
And then, of course,
when the pandemic came,
91
00:05:35,062 --> 00:05:39,244
we realized that there are trade offs
involved when everything is outsourced.
92
00:05:41,024 --> 00:05:46,426
>> Right, and so in that regard, this
news that we heard earlier in the week
93
00:05:46,426 --> 00:05:51,844
that SK Hynix is investing in a new
advanced packaging plant for memory.
94
00:05:53,404 --> 00:05:57,644
Why is it such a big deal for
the US and the Midwest in general?
95
00:05:57,644 --> 00:06:01,301
>> Yeah,
this is really a very big deal for us.
96
00:06:01,301 --> 00:06:05,949
It will create a significant number
of high paying jobs for Hoosiers,
97
00:06:05,949 --> 00:06:09,338
it'll provide opportunities for
our students for
98
00:06:09,338 --> 00:06:13,084
co-ops, internships,
full time experiences.
99
00:06:13,084 --> 00:06:16,812
Not only are they building a fab,
they're building an R and D facility.
100
00:06:16,812 --> 00:06:19,215
So they'll be an intimate
part of the R and
101
00:06:19,215 --> 00:06:23,244
D effort in the US working with
our faculty and graduate students.
102
00:06:23,244 --> 00:06:28,542
So this is an area of the country that
hasn't seen much of this industry.
103
00:06:28,542 --> 00:06:33,349
But I think this is an indication
that with the talent here and
104
00:06:33,349 --> 00:06:38,345
with the capabilities that we have,
this is probably the first
105
00:06:38,345 --> 00:06:43,266
of several announcements that
we're going to see like this.
106
00:06:43,266 --> 00:06:48,408
>> From the crossroads of America,
I guess we're evolving
107
00:06:48,408 --> 00:06:53,879
to the silicon heartland right
here in Indiana, exciting.
108
00:06:53,879 --> 00:06:59,661
Now, the Chips Act is helping bring
some of that critical semiconductor
109
00:06:59,661 --> 00:07:04,321
manufacturing and
packaging supply chain back to the US.
110
00:07:04,321 --> 00:07:07,922
Including the latest news
of course from SK Hynix.
111
00:07:07,922 --> 00:07:13,029
But what are the larger ecosystem
challenges that you see as
112
00:07:13,029 --> 00:07:19,448
we try to bring back semiconductor
manufacturing to the United States?
113
00:07:19,448 --> 00:07:24,272
Workforce, supply chains, labor,
automation, energy, all those aspects.
114
00:07:24,272 --> 00:07:26,184
What else do we need to be thinking about?
115
00:07:26,184 --> 00:07:28,256
>> Well, workforce is the number one.
116
00:07:28,256 --> 00:07:31,024
When you talk to semiconductor
companies CEO's, and
117
00:07:31,024 --> 00:07:34,744
you ask them what's the biggest challenge,
what keeps you up at night?
118
00:07:34,744 --> 00:07:35,824
It's workforce.
119
00:07:35,824 --> 00:07:40,142
We're building factories, where are we
going to find the engineers and
120
00:07:40,142 --> 00:07:44,059
skilled technicians for
these new plants that we're building?
121
00:07:44,059 --> 00:07:49,959
Also securing our overseas supply chain,
we can't do everything in the US.
122
00:07:49,959 --> 00:07:55,135
So there's a lot of work going on
right now, a lot of discussions in
123
00:07:55,135 --> 00:08:00,314
places like Costa Rica, Mexico,
Dominican Republic, Brazil,
124
00:08:00,314 --> 00:08:05,861
even Vietnam in moving the parts of
the supply chain to countries that we
125
00:08:05,861 --> 00:08:10,965
just feel are gonna be dependable and
reliable for the long term.
126
00:08:10,965 --> 00:08:12,244
>> Well, that's fantastic.
127
00:08:13,944 --> 00:08:19,404
Now, Doctor Lundstrom, you have
been instrumental in helping Purdue
128
00:08:19,404 --> 00:08:25,184
play an outsized role in meeting
the semiconductor challenge.
129
00:08:25,184 --> 00:08:29,912
Purdue created the nation's first
comprehensive semiconductor degrees
130
00:08:29,912 --> 00:08:30,593
program.
131
00:08:30,593 --> 00:08:35,892
As a focal point for research, teaching,
fab innovation, collaborations with
132
00:08:35,892 --> 00:08:41,293
the private and public sector, and
most particularly workforce development.
133
00:08:41,293 --> 00:08:44,669
In fact the Washington Post has said that,
quote,
134
00:08:44,669 --> 00:08:49,145
by rapidly expanding chip education,
Purdue is aiming
to graduate
135
00:08:49,145 --> 00:08:54,533
thousands of semiconductor engineers
annually as soon as possible.
136
00:08:54,533 --> 00:08:59,409
Purdue is rolling out new courses and
labs for undergraduates, a new master's
137
00:08:59,409 --> 00:09:03,793
program, and a push to place students
in chip internships, unquote.
138
00:09:04,944 --> 00:09:07,008
That sounds like a lot.
139
00:09:07,008 --> 00:09:10,930
Doctor Lundstrom, what are some of
the highlights of what Purdue is doing,
140
00:09:10,930 --> 00:09:14,930
especially around workforce development
for the semiconductor industry?
141
00:09:14,930 --> 00:09:19,634
>> Okay, well, thank you, Arvind.
142
00:09:19,634 --> 00:09:23,880
This is I believe the nation's
most ambitious effort to develop
143
00:09:23,880 --> 00:09:27,745
the workforce for
this industry that we're reassuring.
144
00:09:27,745 --> 00:09:31,964
And I don't know of any other university
that has made the commitment that
145
00:09:31,964 --> 00:09:35,844
the College of Engineering has made,
and the university has made to
146
00:09:35,844 --> 00:09:40,604
helping the nation address really its
most urgent, immediate challenge.
147
00:09:40,604 --> 00:09:42,304
The program is very special,
148
00:09:42,304 --> 00:09:47,284
we've been teaching semiconductor
courses since the late 1960s.
149
00:09:47,284 --> 00:09:51,292
What's different now is that were not just
teaching them in electrical engineering.
150
00:09:51,292 --> 00:09:57,444
This is a college wide effort,
now its expanding across the university.
151
00:09:57,444 --> 00:10:02,064
There is such a broad need for talent in
this industry, whether you're a materials
152
00:10:02,064 --> 00:10:06,816
engineer or mechanical engineer, an
electrical engineer, a computer scientist,
153
00:10:06,816 --> 00:10:09,341
an industrial engineer, a data scientist.
154
00:10:09,341 --> 00:10:13,381
There are just an enormous number of
exciting career opportunities for
155
00:10:13,381 --> 00:10:14,264
our students.
156
00:10:16,084 --> 00:10:18,884
And that's what's exciting
about this program.
157
00:10:18,884 --> 00:10:23,656
So, there are two particular new
courses that we've launched that
158
00:10:23,656 --> 00:10:28,484
I'm most proud of,
one we call changing the world with chips.
159
00:10:28,484 --> 00:10:33,030
And this is a course where a different
company comes in every week and
160
00:10:33,030 --> 00:10:38,314
talks about what they do, what their
role is in the semiconductor ecosystem.
161
00:10:38,314 --> 00:10:41,505
What kind of career
opportunities there are for
162
00:10:41,505 --> 00:10:46,404
students in their industry,
in their part of the business.
163
00:10:46,404 --> 00:10:49,824
It's directed at first year students,
we maxed out for
164
00:10:49,824 --> 00:10:52,876
two years in a row now
the size of the classroom.
165
00:10:52,876 --> 00:10:55,548
So we're adding a second
session every fall.
166
00:10:55,548 --> 00:10:58,374
So students are really paying
attention.
167
00:10:58,374 --> 00:11:02,395
And if we get students excited
about careers in semiconductors,
168
00:11:02,395 --> 00:11:06,254
what we do next is we offer
a very special summer program.
169
00:11:06,254 --> 00:11:10,741
Now, we've been doing summer research
experiences for students for
170
00:11:10,741 --> 00:11:13,943
20 plus years, and
it's a wonderful program.
171
00:11:13,943 --> 00:11:16,337
It pairs a student with a faculty member,
and
172
00:11:16,337 --> 00:11:19,490
they work with graduate students
in the professor's lab.
173
00:11:19,490 --> 00:11:22,758
This is different,
it's very focused on semiconductors.
174
00:11:22,758 --> 00:11:27,894
It's an intense semiconductor experience,
eight weeks long.
175
00:11:27,894 --> 00:11:32,671
One track is on how to design a chip,
one track is on how to manufacture or
176
00:11:32,671 --> 00:11:33,686
make chips.
177
00:11:33,686 --> 00:11:37,670
There will be a third track on advanced
packaging in the coming summer.
178
00:11:37,670 --> 00:11:41,790
And the idea is to prepare students,
get them excited about considering
179
00:11:41,790 --> 00:11:45,113
this among the many career
opportunities that they have.
180
00:11:45,113 --> 00:11:49,773
And prepare them to do internships with
these companies that are supporting
181
00:11:49,773 --> 00:11:53,454
these students after their sophomore year.
182
00:11:53,454 --> 00:11:58,123
So when we launched this course,
we hoped last summer for
183
00:11:58,123 --> 00:12:02,854
maybe 50 students,
we had more than 500 apply.
184
00:12:02,854 --> 00:12:04,678
And that's been a big question.
185
00:12:04,678 --> 00:12:07,390
With all of the opportunities
that students have,
186
00:12:07,390 --> 00:12:11,782
could we get them excited and interested
in pursuing careers in this industry?
187
00:12:11,782 --> 00:12:14,750
And we're getting all kinds
of signals that we can.
188
00:12:14,750 --> 00:12:19,536
This year, we set an even more ambitious
goal, 100 students in the summer program.
189
00:12:19,536 --> 00:12:24,403
But we had to shut off applications
the first week of February because we had
190
00:12:24,403 --> 00:12:26,760
well over 750 applications.
191
00:12:26,760 --> 00:12:30,728
So students are interested
in what this is all about.
192
00:12:30,728 --> 00:12:33,188
They're excited about
the career opportunities, and
193
00:12:33,188 --> 00:12:35,664
that's really gratifying to see.
194
00:12:35,664 --> 00:12:40,096
>> Well, to our audience,
you heard it right here.
195
00:12:40,096 --> 00:12:43,435
No better place to learn about
semiconductors than here
196
00:12:43,435 --> 00:12:45,564
at Purdue University.
197
00:12:45,564 --> 00:12:47,557
It's so exciting, Doctor Lundstrom,
198
00:12:47,557 --> 00:12:51,435
I feel like I need to get back into this
myself,and take some of these classes.
199
00:12:51,435 --> 00:12:54,204
[LAUGH]
>> And Arvin, just getting back to Purdue
200
00:12:54,204 --> 00:12:57,608
for a while, we're beginning to
see new students coming here.
201
00:12:57,608 --> 00:13:01,738
And they're saying, well, they applied
to several different universities, and
202
00:13:01,738 --> 00:13:03,783
I asked them, why did you come to Purdue?
203
00:13:03,783 --> 00:13:06,461
And they thought, well,
I'm interested in semiconductors,
204
00:13:06,461 --> 00:13:07,772
maybe I should go to Purdue.
205
00:13:07,772 --> 00:13:09,370
So we're getting more and
206
00:13:09,370 --> 00:13:13,484
more of that kind of reputation
as the semiconductor university.
207
00:13:13,484 --> 00:13:15,925
>> How about that?
208
00:13:15,925 --> 00:13:20,484
[COUGH] Doctor Lundstrom,
what's the next frontier?
209
00:13:20,484 --> 00:13:26,887
Once the advantages of advanced packaging
and chiplets have been realized,
210
00:13:26,887 --> 00:13:32,699
and advanced lithography techniques
have met the limits of physics,
211
00:13:32,699 --> 00:13:35,628
what lies beyond 10 years plus?
212
00:13:35,628 --> 00:13:41,115
What will engineers be working
on to roughly meet Moore's law,
213
00:13:41,115 --> 00:13:43,564
a decade or more from today?
214
00:13:44,664 --> 00:13:47,916
Well, that's an interesting question.
215
00:13:47,916 --> 00:13:52,624
Frankly, I think no one can predict
what's going to happen 10 years from now.
216
00:13:52,624 --> 00:13:54,388
And I'll tell you a story,
217
00:13:54,388 --> 00:13:58,568
my first job in the semiconductor
industry was 50 years ago.
218
00:13:58,568 --> 00:14:02,627
In the coming September, it'll be 50 years
that I've been in this industry, and
219
00:14:02,627 --> 00:14:04,128
I worked for Hewlett Packard.
220
00:14:04,128 --> 00:14:07,806
At that time,
every company manufactured its own chips,
221
00:14:07,806 --> 00:14:09,917
and we were making our own chips.
222
00:14:09,917 --> 00:14:14,292
Hewlett Packard had a crash program
underway at that time because they saw
223
00:14:14,292 --> 00:14:16,279
the end of optical lithography.
224
00:14:16,279 --> 00:14:18,465
And he said, it's just basic physics,
225
00:14:18,465 --> 00:14:22,146
you can't make a pattern smaller
than the wavelength of the light.
226
00:14:22,146 --> 00:14:27,094
Well, actually, you tell an engineer that,
and that's a challenge to them.
227
00:14:28,234 --> 00:14:32,626
We actually learned to make several
different ways to make patterns smaller
228
00:14:32,626 --> 00:14:36,266
than the wavelength of the light
used to expose the pattern.
229
00:14:36,266 --> 00:14:41,123
And then we manufactured ultra
sophisticated tools that had shorter and
230
00:14:41,123 --> 00:14:42,898
shorter wavelengths.
231
00:14:42,898 --> 00:14:47,749
And last spring break we took
22 Purdue students to Europe on
232
00:14:47,749 --> 00:14:51,634
a semiconductor study abroad experience.
233
00:14:51,634 --> 00:14:56,964
We spent a day at ASML, who makes this
very sophisticated lithography tool,
234
00:14:56,964 --> 00:15:03,226
and Arvind, they have a 20-year roadmap,
so this isn't going to end anytime
soon.
235
00:15:03,226 --> 00:15:07,582
And frankly,
I think as long as I'm on this planet,
236
00:15:07,582 --> 00:15:11,549
we'll be living in this
CMOS-centric world.
237
00:15:11,549 --> 00:15:14,788
>> [LAUGH] That is fantastic.
238
00:15:14,788 --> 00:15:18,154
You mentioned your 50-year career,
239
00:15:18,154 --> 00:15:22,807
remarkable career, if I may add,
in semiconductors.
240
00:15:22,807 --> 00:15:27,713
Can you tell us a little bit about how
you got interested in semiconductors
241
00:15:27,713 --> 00:15:29,068
in the first place?
242
00:15:29,068 --> 00:15:34,539
What was your personal journey to
deciding this is what you wanted to do,
243
00:15:34,539 --> 00:15:37,512
and why are you so passionate about it?
244
00:15:37,512 --> 00:15:41,722
>> Well, like many students,
sometimes you just take a course, and
245
00:15:41,722 --> 00:15:46,777
it just resonates with you, and you think,
wow, this is cool, this is for me.
246
00:15:46,777 --> 00:15:51,269
As a kid in high school,
I was a ham radio operator, and
247
00:15:51,269 --> 00:15:56,463
I tinkered around with electronics,
I knew there were chips.
248
00:15:56,463 --> 00:15:59,143
At that time,
chips had maybe tens of transistors, or
249
00:15:59,143 --> 00:16:02,423
maybe 100 in the most sophisticated one,
I knew they existed.
250
00:16:02,423 --> 00:16:05,087
And then when I was
a junior undergraduate,
251
00:16:05,087 --> 00:16:09,263
I took a course from someone who had
just left industry, had spent his
252
00:16:09,263 --> 00:16:13,306
career setting up semiconductor
factories in different places.
253
00:16:13,306 --> 00:16:17,622
He had done one in Motorola in Arizona,
at that time.
254
00:16:17,622 --> 00:16:21,433
He came back, and he was teaching
this course, and he said, class,
255
00:16:21,433 --> 00:16:24,270
I've been out of the industry for
two years now.
256
00:16:24,270 --> 00:16:27,562
It moves so fast that I cannot
tell you what it's like.
257
00:16:27,562 --> 00:16:30,050
And I thought, okay,
that's really exciting.
258
00:16:30,050 --> 00:16:33,514
He says, we can only talk
about some general principles.
259
00:16:33,514 --> 00:16:38,538
We read Gordon Moore's paper,
we studied the yield mechanisms.
260
00:16:38,538 --> 00:16:42,332
We talked about how you decide how
many transistors you put on a chip
261
00:16:42,332 --> 00:16:46,842
until the yield drops so much that it's
better to put them on several chips.
262
00:16:46,842 --> 00:16:50,194
And I thought, this is really cool,
this is interesting.
263
00:16:50,194 --> 00:16:53,858
So I stayed one more year and
got a master's degree, worked in
264
00:16:53,858 --> 00:16:59,334
the semiconductor lab at University of
Minnesota, another land grant university.
265
00:16:59,334 --> 00:17:04,062
And then my very first job was in
semiconductor process development for
266
00:17:04,062 --> 00:17:05,374
Hewlett Packard.
267
00:17:05,374 --> 00:17:09,157
We were building desktop
computers at that time, and
268
00:17:09,157 --> 00:17:12,654
I was working on a five
micrometer process.
269
00:17:12,654 --> 00:17:14,778
This was really pushing the state of art,
270
00:17:14,778 --> 00:17:17,254
this was really difficult
to make this work.
271
00:17:17,254 --> 00:17:20,718
The previous process was
a seven micrometer process.
272
00:17:20,718 --> 00:17:23,406
Five is seven divided by
the square root of two, so
273
00:17:23,406 --> 00:17:25,402
the area of the transistor is half.
274
00:17:25,402 --> 00:17:28,802
We could put twice as many on a chip,
we were on Moore's law.
275
00:17:28,802 --> 00:17:32,814
We weren't really thinking about it, and
we had no idea how long it would continue.
276
00:17:32,814 --> 00:17:36,931
And I've been in the field ever since.
277
00:17:36,931 --> 00:17:43,354
I spent a few years at Hewlett Packard,
came back to Purdue to do my PhD,
278
00:17:43,354 --> 00:17:48,292
and I've been working in
semiconductors ever since.
279
00:17:48,292 --> 00:17:50,147
It's an exciting field,
280
00:17:50,147 --> 00:17:55,650
it's the most sophisticated technology
that human beings manufacture.
281
00:17:55,650 --> 00:17:58,376
It's foundational to
the life as we know it,
282
00:17:58,376 --> 00:18:01,474
would not exist without this technology.
283
00:18:01,474 --> 00:18:06,489
And to have had an opportunity to
spend my career working in this field,
284
00:18:06,489 --> 00:18:11,818
and teaching other students how to
enter this field and contribute to it.
285
00:18:11,818 --> 00:18:13,334
I've just been very lucky.
286
00:18:14,824 --> 00:18:21,471
>> Doctor Mark Lundstrom, leader,
innovator, and teacher, extraordinaire.
287
00:18:21,471 --> 00:18:24,496
The Mark Lundstrom Research Group
explores the physics and
288
00:18:24,496 --> 00:18:27,104
technology of nano electronics devices.
289
00:18:27,104 --> 00:18:32,342
As early as the 1990s, he founded the
Purdue University network computing hubs,
290
00:18:32,342 --> 00:18:36,005
which provided online simulation
services for research and
291
00:18:36,005 --> 00:18:38,744
education in micro and nano electronics.
292
00:18:38,744 --> 00:18:42,275
His work there led to
the creation of the Network for
293
00:18:42,275 --> 00:18:46,138
Computational Nanotechnology and
the nanohub.org,
294
00:18:46,138 --> 00:18:50,964
which today serve the nanotechnology
community worldwide.
295
00:18:50,964 --> 00:18:54,764
Check out Purdue and
semiconductors at the website,
296
00:18:54,764 --> 00:18:58,320
engineering.purdue.edu/semiconductors.
297
00:18:58,320 --> 00:19:04,546
This is Arvind Raman signing
off this episode of Engenuity.
298
00:19:04,546 --> 00:19:08,383
[MUSIC]
299
00:19:08,383 --> 00:19:12,658
>> This has been Engenuity,
a Boilermaker Engineering podcast.
300
00:19:12,658 --> 00:19:16,124
For more information on some of
the issues discussed in this podcast,
301
00:19:16,124 --> 00:19:20,194
go to the engineering homepage, and click
on the podcast icon on the upper right.
302
00:19:20,194 --> 00:19:20,694
[MUSIC]