[CS61C FA20] Weekly Lecture 12.LIVE - SIMD — Transcript
Full transcript
- 0:00computer all right ladies and gentlemen
- 0:03welcome to cs61c
- 0:05once a week lectures live session on
- 0:07parallelism
- 0:09great to have you here great to have you
- 0:11here good stuff
- 0:13hope you boy we're as rapped as i were
- 0:16was
- 0:16uh last week with cnn i don't think i've
- 0:19watched more hours of cnn
- 0:21in my life than i did last week was
- 0:23everybody were you
- 0:24locked into that i had it on on lockdown
- 0:26over here i had my
- 0:28second monitor always we
- 0:31i mean i don't watch television but yes
- 0:33i spent too many hours and you know
- 0:35there was an impact that i posted uh you
- 0:38know today's videos for the lecture
- 0:40last night at like 2 a.m because i had
- 0:42to
- 0:45you were on last week that's true when
- 0:47you're on it's hard to have something
- 0:48taken i was obviously yeah i had a
- 0:50little chance that's right
- 0:51wonderful all right so let's let's go
- 0:53for what we're talking about today
- 0:54today's agenda
- 0:55again welcome to week 12 everybody we
- 0:56have some wonderful guests from apple
- 0:58jared dave and anand and we'll introduce
- 1:01them then we're going to talk about
- 1:02computing in the news i've got a
- 1:03delightfully humorous competing the news
- 1:05to share with you
- 1:06uh a schedule there's no new policy
- 1:08updates just just fyi
- 1:09you missed that bullet is usually there
- 1:11we're missing that no policy updates and
- 1:12then we'll have an ask me anything to
- 1:14talk about
- 1:14you know anything else you have folks i
- 1:16think things are on students minds
- 1:18so most of the time most of the time is
- 1:22going to go
- 1:22to um to our chat with applefox
- 1:26right we have um uh dave conroy jared
- 1:29zerby
- 1:30and anand uh with us we would like to
- 1:33ask them
- 1:33you know how they ended up
- 1:37over um uh in in
- 1:40in in this part of their careers but
- 1:43then uh us
- 1:44after that we're gonna ask them for
- 1:46advice um on
- 1:48internships because we have heard
- 1:50earlier in the semester that apple
- 1:52is interested in interns and full-time
- 1:54employees
- 1:55um so we're going to try to spend
- 1:58uh most of the time um on that
- 2:02but let's start with the introductions
- 2:04um and um
- 2:05so maybe we can start with the dave
- 2:08because
- 2:08uh i think people know jared already
- 2:13dave how did you end up were you did you
- 2:16decide you know when you when you
- 2:18started kindergarten you know
- 2:19one day i'm gonna work for apple
- 2:23and tell us also what your role for
- 2:24apple is now people may have not heard
- 2:26that
- 2:26sure i'm dave conroy um i'm a i'm a
- 2:29fellow
- 2:29at at apple uh a fellow is a great goal
- 2:33you uh you get to have all the uh the uh
- 2:36benefits of being a vice president
- 2:38without having to do all the management
- 2:39stuff of being a vice president
- 2:41so it's great for a technical person
- 2:42like me um i've been at apple for 17
- 2:45years
- 2:46um and uh before that i did a short
- 2:49stint at
- 2:50a consumer electronics startup that got
- 2:51acquired by microsoft and before that i
- 2:53did 20 years of digital equipment
- 2:56who many of you don't know who they are
- 2:59but they were the people who invented
- 3:00mini computers on the east coast that
- 3:02they were like a force
- 3:03of their era were you our east coast
- 3:05person at the time
- 3:06dave yeah i was in maynard i was in
- 3:08canada i was in maynard and hudson
- 3:10and then i moved out to palo alto and
- 3:12worked in bob taylor's lab
- 3:14at the end right so i know all those
- 3:16xerox park guys pretty well sure sure
- 3:19my path to apple is when i was at
- 3:22digital
- 3:22apple was building this thing called
- 3:24newton that some of you will remember
- 3:27and i was building a microprocessor
- 3:29called strongarm which most people
- 3:30didn't know
- 3:32um went into a second generation
- 3:35newton that never shipped and uh
- 3:40and so it all you know we we were
- 3:42building this thing and then that newton
- 3:43got cancelled
- 3:44and uh and it all kind of disappeared
- 3:48and then years later i got this call
- 3:49from this guy named mike colbert out of
- 3:51the blue
- 3:52uh who was at apple who knew me from
- 3:54like back when we worked together like a
- 3:56decade before
- 3:57on this newton and said you should come
- 3:59to apple it'll be fun
- 4:01and uh and of course this is 17 years
- 4:03ago which means if you're going to come
- 4:05to apple
- 4:06you have to decide how comfortable you
- 4:08are going to a place that's going to go
- 4:10bankrupt in the next six months and
- 4:12you're going to have to find another job
- 4:14but i liked the people and so i said
- 4:16what the hell i could find a job in six
- 4:18months if y'all goes
- 4:19the ball goes to crap and uh it turned
- 4:21out it worked out kind of okay
- 4:23um i've worked on a whole bunch of
- 4:25things that at
- 4:27apple i started off working on the intel
- 4:29transition which turned out was the
- 4:31reason
- 4:32um that mike wanted me to come
- 4:35but i've worked on silicon for the
- 4:37phones i've worked with i work
- 4:39collaborated with intel on the chips
- 4:40that went into the portables i've worked
- 4:42on i o standards i've worked on all
- 4:44sorts of things
- 4:46so that's me oh one one thing
- 4:49clarification here how many apple
- 4:51fellows are
- 4:51there i mean there is like 5 000 or like
- 4:55less two or three if you count phil
- 4:58schiller three three i think
- 4:59yeah phil showed him three if you count
- 5:01our marketing fellow
- 5:04all right we call that not that many we
- 5:07call that rare air
- 5:08well done dave all right nice
- 5:14um jared uh how did you end up in that
- 5:17lab behind you know that's behind you
- 5:19all right it's it's really great because
- 5:21i got i got a two vice presidents behind
- 5:24me
- 5:24just in case i need you know backup so i
- 5:26can i can i can you know yell back to
- 5:28them at any time
- 5:29um no i i mean uh so my path to apple
- 5:32was
- 5:33i when i started uh college i thought i
- 5:35was gonna be an aeronautical engineer
- 5:38and then i took my first aeronautical
- 5:39engineering class and i said this is
- 5:41ridiculous
- 5:42these people have no idea what's going
- 5:43on i mean there's
- 5:45you know it's just fudge factor after
- 5:47fudge factor there's no science here at
- 5:48all it's just a you know a complete
- 5:50slapdash thing um and then
- 5:53uh you know i started i had always loved
- 5:56tinkering around with computers so
- 5:57um so i started looking at double e and
- 6:00this was uh back in the
- 6:02uh early to mid 80s uh so
- 6:06um i took a class uh
- 6:09across the bay from where you guys are
- 6:12and
- 6:13got to design and build my own chip
- 6:17and i just thought that was the best
- 6:19thing i had ever
- 6:20uh i had ever seen you know the fact
- 6:23that i wasn't going to be doing some
- 6:25problem set that some professor had
- 6:26dreamed up at
- 6:27two or three in the morning and you know
- 6:30and i
- 6:30i could actually build something of my
- 6:32own and have it
- 6:34you know be something different um and i
- 6:37just found the
- 6:38ability to express uh uh
- 6:41you know ideas in silicon to be just
- 6:44fascinating
- 6:44and so that was 33 years ago
- 6:48uh after that i uh
- 6:51i when i graduated and i would like to
- 6:54impress on people the
- 6:55impact of when you are interviewing
- 6:59uh and you bring to the interview a die
- 7:02photo a plot and a schematic of the chip
- 7:06that you made
- 7:07it has a different kind of impression on
- 7:09on people then rather than like you know
- 7:11digging around in your
- 7:12transcript for a grade you know or
- 7:14showing someone a problem set
- 7:16because for somebody in industry to see
- 7:18a student that's actually built
- 7:20something and understands how to bring
- 7:21it up and
- 7:22what the well you know where the the
- 7:24bugs are and stuff like that it's it's
- 7:25different
- 7:26it's a different kind of interview let
- 7:27me just say that um so
- 7:29so uh it usually goes well for the
- 7:32student it's what we're trying to say
- 7:33here
- 7:34it goes regularly very well for this it
- 7:36goes very well
- 7:37uh it opens a lot of doors so there's a
- 7:39glass
- 7:40there's a class uh 194 that we're
- 7:43helping uh
- 7:44you know sponsor at berkeley for people
- 7:46who want to tape out their own chips
- 7:49uh it's a lot of work but it's a lot of
- 7:51fun a lot of these classes do logistim
- 7:53diagrams
- 7:54count
- 7:54[Laughter]
- 7:58that's good it helps but it's but it's
- 8:00not the same as actually building
- 8:02something
- 8:03um so uh
- 8:07don't tell them
- 8:10i went to a company called uh vlsi
- 8:12technology then i went to mips computer
- 8:14systems i was designing floating point
- 8:15units
- 8:16and i went to rambus for 21 years so
- 8:18when i look at dave i like dave's a deck
- 8:20guy
- 8:20you know so i think people still look at
- 8:22me and thinking that i'm a rambus guy
- 8:24um because that's where you know you you
- 8:26put in a lot of time there i've been in
- 8:28apple for seven years
- 8:30um i'm fortunate enough to work for the
- 8:32other apple fellow
- 8:33that's that's that's technical um not
- 8:36phil schiller
- 8:37and um and uh i work in a team called
- 8:41exploratory design
- 8:42which is uh really a lot of fun because
- 8:44uh we get to do a lot of
- 8:46uh crazy things across lots of lots of
- 8:48different
- 8:50spans and uh before we get into the q a
- 8:53i just want to make
- 8:54two points uh number one no we can't
- 8:56talk about anything
- 8:58about uh what's gonna be announced
- 8:59tomorrow so if you can just take that
- 9:01off of your
- 9:02questions list um because uh you know
- 9:04we're not gonna get an on
- 9:06fire uh in the process of of this
- 9:08meeting
- 9:09uh uh and then uh you know number two
- 9:13uh yes apple has internships yes apple
- 9:15has internships for undergrads and
- 9:17hardware
- 9:17um in fact the largest organization in
- 9:20hardware technology which has the most
- 9:22internships in the in the hundreds
- 9:24uh uh for for this coming summer
- 9:27has a charter too and they are
- 9:29successful at
- 9:30hiring 50 undergrads for internships
- 9:34so um so just recalibrate yourself on
- 9:38that
- 9:40all right and and maybe you can
- 9:43introduce yourselves
- 9:44you know did was your path to apple the
- 9:47most
- 9:48straightforward these guys took a little
- 9:49bit of a windy path
- 9:52yeah and what's your what's your role in
- 9:53apple as well yeah share that
- 9:55uh sure so um i work in uh
- 9:58a group at apple called platform
- 9:59architecture uh i manage a few teams
- 10:01that are focused on at a high level you
- 10:03can think about it as power performance
- 10:05and prototyping um so we study how fast
- 10:08things are how much power they draw
- 10:10and try and kind of create new things
- 10:12that that we might turn into
- 10:13products or technologies in the future
- 10:16um my path sample was a little
- 10:18unusual um so i've only had two jobs my
- 10:20whole life my first job uh was company i
- 10:22started when i was 14.
- 10:24um it was a deep technology publication
- 10:26called a non-tech so my first name plus
- 10:29tech at the end you can kind of guess
- 10:30what that was that was about
- 10:32um so i started that back in april of 97
- 10:35um
- 10:36kind of showed up at the just before the
- 10:38dot-com boom or as the dot-com boom was
- 10:40happening
- 10:41um and so i i rode that wave up i never
- 10:44took any vc or anything like that this
- 10:46is entirely uh uh homegrown
- 10:48um so i i rode that wave up i lived
- 10:51through the dot com bust i
- 10:52lived through the kind of uh rebuilding
- 10:55web 2.0
- 10:57um and then lived through the great
- 10:58recession managing the company as well
- 11:00and then again we rebuilt and continued
- 11:02to grow um
- 11:04and as i was kind of you know entering
- 11:06into my 30s i've been kind of doing the
- 11:08same thing for a very long time my goal
- 11:09always was to like you know
- 11:11at some point uh kind of take a step
- 11:13back
- 11:14i always said that you know every big
- 11:16transition my company had
- 11:18uh we always grew a lot i ended up on a
- 11:20plane a lot more and i ended up with
- 11:21less hair
- 11:22uh and so you know we were in like kind
- 11:24of pivot number four
- 11:25over this kind of 18 year time span and
- 11:27i said look it's time for me to slow
- 11:29down a little bit
- 11:30and the plan was to always you know kind
- 11:31of exit and and go stay with the parent
- 11:34company for a while and help them build
- 11:35other verticals similar to
- 11:37what i had built um i always believed
- 11:39that kind of my core innovation there
- 11:40wasn't necessarily
- 11:42how we approached analyzing computer and
- 11:44mobile hardware but our approach to
- 11:46journalism
- 11:47and so i wanted to go and kind of uh
- 11:49duplicate that across a bunch of
- 11:51different verticals
- 11:52around the time i was having these kind
- 11:54of thoughts and and getting the company
- 11:56ready to came and take that next step
- 11:57uh apple approached uh with an interest
- 12:00in hiring uh someone who worked for me
- 12:02um which was not unusual we you know
- 12:04were a good pipeline for
- 12:05uh people who were students they would
- 12:07get a lot of industry experience working
- 12:09for us
- 12:10and then oftentimes they they you know
- 12:11either stuck with us for a long time or
- 12:13they would go work in industry
- 12:14and kind of give you a leg up over the
- 12:16competition um so they approached me
- 12:18trying to hire one of my folks again not
- 12:20unusual
- 12:21it was a connection that we enabled and
- 12:22helped make and in the conversation they
- 12:24said hey you know be
- 12:25kind of interesting if you join um and i
- 12:28had never thought about working for any
- 12:30of the companies that we covered
- 12:32um i think most of them by the time
- 12:36product got to us you know we started as
- 12:38a pc hardware and technology review site
- 12:40eventually expanded into consumer
- 12:42electronics servers
- 12:43chips all of that stuff uh i always felt
- 12:46that like you know
- 12:47i could see the politics just from the
- 12:50outside
- 12:51uh that meant that i didn't want to work
- 12:52at most of these companies the only one
- 12:54that you know kind of
- 12:55seemed okay to me was apple um and so i
- 12:58entertained the thought you know i had a
- 13:00conversation with uh the person who
- 13:03became my boss
- 13:04and after that initial conversation i
- 13:05was like yeah this makes no sense but
- 13:07after a bunch of these conversations uh
- 13:09it kind of opened me up to the
- 13:11ability to influence uh and and make an
- 13:14impact on
- 13:15uh just honestly one of the greatest
- 13:17companies in the world uh and then i
- 13:19always figured hey the bar was low
- 13:20enough in journalism that a 14 year old
- 13:22kid that knew nothing could
- 13:23kind of do well in it so if i ever
- 13:24wanted to go back i could i could always
- 13:26do that
- 13:27um and so i've been at apple for a bit
- 13:29over six years now
- 13:30uh and yeah that's my story
- 13:33know th this is really thank you this is
- 13:35really funny um
- 13:36do you know how many times dan and i
- 13:38when we go for this computing in the
- 13:40news uh
- 13:41you know grab stuff from alan we pull it
- 13:44every week it's great
- 13:46you know
- 13:54but this is funny i mean most students
- 13:56do not know that there was a 14 year old
- 13:58kid behind that website
- 14:02[Music]
- 14:03uh when i first started i remember the
- 14:06day someone posted a comment that like
- 14:08hey this kid goes to high school
- 14:10and i thought i was like you're outed
- 14:13you're outed look at that for three
- 14:16months and then everyone thought my dad
- 14:17wrote everything and i was like that's
- 14:19hilarious but uh yeah that's
- 14:20another really good that's brilliant
- 14:23yeah that's great
- 14:24all right um and there are a couple of
- 14:27questions that came in guys feel free
- 14:28asking questions says
- 14:30did any of you get to meet steve wozniak
- 14:33or steve jobs
- 14:36well by the time i showed up
- 14:40you know steve wozniak was long gone um
- 14:43i i would see him occasionally at
- 14:45at launches because he always got to sit
- 14:47in the front row because he always got
- 14:49an invitation
- 14:51i met steve a couple of times
- 14:56most of the stories about steve is true
- 14:58and that means that
- 15:00lots of people spend a lot of time
- 15:02trying to arrange that they perhaps
- 15:03didn't have to be in a meeting with
- 15:05steve
- 15:06um and uh this mike culbert guy that i
- 15:09talked about before
- 15:10uh he and steve were uh were were
- 15:14very very trusteds and so most of the
- 15:17time i could get anything i wanted to
- 15:19see
- 15:19simply by telling it to mike and then
- 15:21mike would tell it to steve and if steve
- 15:23went ra
- 15:24he put bra
- 15:27which was a win
- 15:32all right um
- 15:35there is another question over here
- 15:37should we tell them or should we keep
- 15:38the secret for another day we are not
- 15:40done
- 15:40grading and checking everything um i
- 15:43feel like keeping a surprise
- 15:44i feel like we keep it a secret i don't
- 15:46know you know what's the price
- 15:48for project three and i will find out in
- 15:51a few days
- 15:52you'll find out
- 15:55[Laughter]
- 16:03they said it was shipped to it's
- 16:06incorrect hole somewhere in shipping and
- 16:07receiving
- 16:08we're gonna have a ton of fun getting to
- 16:10that
- 16:13just keep the make sure you you keep
- 16:15feeding it and cleaning it because if
- 16:16otherwise you know
- 16:17[Laughter]
- 16:20you dribble something
- 16:23yes don't shake the box um
- 16:27people ask you is it their own cpu is it
- 16:29is it a custom apple cpu
- 16:31yeah they're they're going they're going
- 16:33it's the a15
- 16:34it's the a15
- 16:38that's funny yeah all right so
- 16:41i mean one of the key things that we
- 16:43wanted to talk about here are uh
- 16:45to elaborate a bit on what jared said um
- 16:48about these internships uh right i mean
- 16:51lots of our students
- 16:52uh find a place in in internships in um
- 16:56in software as undergrads and they they
- 16:59enjoy them
- 17:00um and you know a good number goes to
- 17:03apple and does uh
- 17:04a software internship um but often there
- 17:07is this um
- 17:09idea that you can't get the you know the
- 17:11the you can get an internship in
- 17:13hardware
- 17:14uh now you're telling us that is uh um
- 17:17you know everybody's wrong about that
- 17:20and
- 17:20uh um can you tell us a little bit more
- 17:23what's the secret there what what do you
- 17:25need to
- 17:25to to know you know can you do it with
- 17:28uh
- 17:2861c type of knowledge um
- 17:32i mean do you need to pass by like dave
- 17:34interview
- 17:35you know the the other fellow uh
- 17:37activist belatas
- 17:38you know do they need to grill you for a
- 17:40little bit uh and see how you do
- 17:42i i sure hope not no i mean i you know
- 17:47no uh most most people are coming for
- 17:49internships will not uh
- 17:51would not pass the uh the deal the the
- 17:54bill and bill and dave uh you know
- 17:56uh phalanx i think i don't think i would
- 17:59pass a bill
- 18:02that's true yeah you know these are the
- 18:03kinds of things that we don't want to
- 18:04ask too of any questions about but
- 18:06um the i mean 61c is absolutely
- 18:10you know getting into into understanding
- 18:12the way that machines work at the very
- 18:14you know deep level on a register level
- 18:16um is is absolutely very important
- 18:18there's a there are a lot of openings
- 18:20in uh in verification
- 18:24uh is really applicable because you
- 18:26really didn't deeply understand
- 18:28where the machine's going um there are a
- 18:31fair amount of
- 18:32uh openings in embedded software that's
- 18:34that that is
- 18:35that are that are immediately applicable
- 18:37with a 61c background
- 18:39um but the more that you can branch out
- 18:42and the more that you can do and show in
- 18:43terms of your projects and and project
- 18:45background
- 18:45interests in my view those are the
- 18:47things that stand out
- 18:48and on a resume or are projects uh
- 18:52you know how you do in labs how you
- 18:53doing projects talking
- 18:55more about what it is that you've that
- 18:57you've made that makes a big difference
- 18:58i don't know um you know david and on
- 19:00what you guys think about um
- 19:02students and software and hardware and
- 19:04how how
- 19:05you see them coming in and what you see
- 19:07is being successful
- 19:10i mean i'm a super bowl i'm as you know
- 19:12i'm a super believer in internships
- 19:14um i you know it's
- 19:19when you're sort of learning this this
- 19:22trade for lack of a better word
- 19:24you know a really really important part
- 19:27of it
- 19:27is to have a really really solid
- 19:30foundation
- 19:31on on the theory and the next click
- 19:33above that and academia does a
- 19:35spectacular job of doing that
- 19:37but you know a part of learning a trade
- 19:40is just learning how to get your hands
- 19:41dirty and the millions of little jobs
- 19:43that have to be done
- 19:45and you know as someone who spent their
- 19:47whole life sort of building computers
- 19:48and shipping them to people
- 19:50it's amazing how much of the work is
- 19:52just work
- 19:54as opposed to um
- 19:57as opposed to like you know unbelievably
- 20:00fancy remarkable earth-shattering work
- 20:02there's just a lot of stuff that has to
- 20:04be done
- 20:05um and and it's important stuff it's not
- 20:08you got to do it you know you have to
- 20:11kind of slog your way through it
- 20:12and the good news for that is it means
- 20:15that
- 20:16you know for someone who's sort of at
- 20:18the beginning of
- 20:19their career like an undergraduate um in
- 20:22in ecs right that you can come to a
- 20:26place like apple
- 20:27and do some stuff that you're that
- 20:31you're perfectly capable of doing with a
- 20:33little bit of the help
- 20:34from the people who are already there
- 20:37and it's important
- 20:38right that that you know that this this
- 20:41this
- 20:42stuff that you're doing and you're kind
- 20:44of you know learning
- 20:45you know you're learning how to set up
- 20:48the thing for this cad tool and learning
- 20:50how awful this cad tool is so that you
- 20:52can
- 20:53be part of the people who know how awful
- 20:54this cad tool is right that if this
- 20:56doesn't happen
- 20:57there is no phone next year right it's
- 21:01right
- 21:01it's it's really good and you come out
- 21:03of it you get to make a contribution to
- 21:05stuff that that you get to see in the
- 21:06world and you get to learn just
- 21:08you know a little bit of the hands dirty
- 21:10part of the trade
- 21:12right i've i'm a i'm from my background
- 21:15i'm a waterloo guy and waterloo is a
- 21:17place that believes in cooperative
- 21:18education you can't get an engineering
- 21:20degree without
- 21:21doing it um and i had the good fortune
- 21:25to
- 21:26you know at the beginning to get tied in
- 21:27with a bunch of guys in this little
- 21:29engineering
- 21:30company who you know remembered building
- 21:32computers out of vacuum tubes and i
- 21:33learned a
- 21:34billion things from them right it was
- 21:36just really really great
- 21:40that's awesome i can tell you yeah go
- 21:42ahead
- 21:43um for me and i think this is true
- 21:46whether it's an internship or you know
- 21:49someone that we're just trying to hire
- 21:50who's out of college
- 21:52um i think technical depth is really
- 21:54important you know be a master of some
- 21:56domain have
- 21:57a lot of skill that you can demonstrate
- 21:58in some area but the thing that i often
- 22:00find is missing
- 22:01is passion right like you know we will
- 22:03interview for a position for a year
- 22:06and find one person out of all the
- 22:09interviews that year that
- 22:10really cares about the stuff they really
- 22:12they have passion for it they're really
- 22:14enthusiastic about it
- 22:15um i think that's a you know if if you
- 22:18have the right intersection of
- 22:19technical depth uh in something whatever
- 22:22your specialty is some technical depth
- 22:24and passion you immediately stand out to
- 22:27me
- 22:27um and it's such a hard thing to find um
- 22:30both amongst
- 22:31uh i think new college grads people who
- 22:34are in industry or even interns but if
- 22:36we find someone like that we
- 22:37i mean those people are the people that
- 22:39get routed to me and they're like you
- 22:40need to go talk to this person
- 22:41um and so i think that i i can't uh uh
- 22:45overstate enough the the value that i
- 22:47think that brings
- 22:49so what do these guys need to put in the
- 22:51resumes for to to
- 22:53make that bar i mean there is a lot of
- 22:55them that we
- 22:58i mean we have i mean what's yeah to
- 23:00that point i mean to add on to
- 23:02what uh anand and dave said i mean i
- 23:04think a lot of it comes down to the
- 23:06person
- 23:06right it really comes down to the
- 23:08individual having the the tenacity and
- 23:10the follow through you know and dave
- 23:11said a lot of is just work that means
- 23:13that what you do
- 23:16uh to bring to you know what you do to
- 23:19differentiate the work that you do
- 23:22you know yeah are you doing quality work
- 23:25are you bringing it you know on time
- 23:26are you are you are you making it happen
- 23:28are you supporting the people around you
- 23:30are you
- 23:31you know being a good team player are
- 23:32you thinking not just about your block
- 23:34what are you thinking about other
- 23:35people's blocks
- 23:36are you studying the code around you are
- 23:39you are you
- 23:40learning beyond what it is that you just
- 23:42need to do
- 23:43there's a great uh book about bell labs
- 23:45and and i think
- 23:47one of the leaders of bellaps said you
- 23:48know what you do um
- 23:51you know in the in the eight hours per
- 23:53day um
- 23:54that pays for your that that justifies
- 23:56your salary and your you know your your
- 23:58medical
- 23:59and uh what you do in the other 16 hours
- 24:02a day and the weekends
- 24:04that's your bonus right there right so
- 24:06you know it's it's thinking about
- 24:08it's not and it's not just like pouring
- 24:10the coals on it it's
- 24:12it's the it's really you know thinking
- 24:15deeply about these things and spending
- 24:17some time and getting kind of getting
- 24:19hooked on
- 24:19on it what it is the uh you know the the
- 24:23you know the joy of it right you know
- 24:25and if you if you're not
- 24:26if you're not loving it then it's not
- 24:28going to come out right so i mean do
- 24:30do the thing that you it's a little bit
- 24:32of a trope but
- 24:33i mean do the thing that you love and
- 24:35and it will show through
- 24:36and and one of the one of the one of my
- 24:39my favorite stories about when i first
- 24:41started at apple was
- 24:42my wife asked me you know so after her
- 24:44first week she said well how's it going
- 24:46and i said
- 24:47it's good but i'm hungry all the time i
- 24:50don't know why i'm hungry all the time
- 24:51and i and it didn't take me long to
- 24:53figure out it was because i would
- 24:55look down get down and start looking at
- 24:58a problem
- 24:58and then start looking at what different
- 25:00people had done and then before i looked
- 25:01up it would be like three or four hours
- 25:03would have passed you know and
- 25:04and it'd be like lunch flew by and i was
- 25:06just like oh yeah this is cool this is
- 25:09really cool
- 25:10this is really really cool you know so
- 25:12it
- 25:13it there are a lot of people who just
- 25:15love to nerd out on what it is that they
- 25:16do at apple and i think that
- 25:18um that definitely is a key
- 25:20characteristic
- 25:21so so i'll ask some questions over here
- 25:25on behalf of the class and then after
- 25:26that you know we got a
- 25:28few q and a uh questions that are coming
- 25:30in um
- 25:33one thing that from my interaction with
- 25:35the students in the class what they've
- 25:36figured out um so far that they seem to
- 25:39be every day
- 25:40more and more comfortable with
- 25:41performance programming embedded
- 25:43programming how to fit things into
- 25:45tight spaces and get maximum performance
- 25:47out of that
- 25:49and they got a good feel for hardware
- 25:51but are
- 25:52one step away from you know in 61c
- 25:56one step away from actually building
- 25:59real hardware
- 25:59uh past the largest scene what's your
- 26:02advice for them i mean
- 26:03first you know do you hire people who
- 26:04need to pack extremely tight
- 26:07matrix multiplication loop and you know
- 26:09run tight neural net
- 26:10into a little embedded space that is out
- 26:13there
- 26:14um and once when they master that you
- 26:16know how do they go
- 26:17and build their own neural net hardware
- 26:20or something else exciting or maybe
- 26:22verify it or
- 26:24and so on i think i i can at least speak
- 26:27from my orgs there
- 26:28there are people i employ who they
- 26:32uh even though we exist in a hardware or
- 26:34their strength is that
- 26:36being in that intersection between
- 26:38really understanding how code runs on
- 26:40on uh on bare metal and understanding
- 26:43the architecture
- 26:44and so they never need to ever build a
- 26:46chip or or verify chip or do anything
- 26:48like that
- 26:48but understanding and being able to
- 26:50exist in that kind of uh
- 26:52gap between the two is uh it's very rare
- 26:55and very important
- 26:56um and so people in that space with a
- 26:58lot of expertise on how to make code
- 27:00run very quickly uh on the underlying
- 27:02architecture
- 27:03that's that's a very important skill to
- 27:06have
- 27:09sounds good and how do you make the next
- 27:11step
- 27:14uh what do you mean by the next maybe
- 27:16others can say yeah how do you actually
- 27:18go
- 27:18you know towards building your own um
- 27:23hardware i mean how do you you know get
- 27:25comfortable interviewing for a harder
- 27:27position
- 27:28for example maybe perhaps a question for
- 27:31jared
- 27:32jared would be better equipped for that
- 27:34um getting comfortable interviewing for
- 27:36a hardware position i
- 27:37um i mean i think there are two parts of
- 27:39it one is that you need to
- 27:42um you need to
- 27:46you need to to come into it with some
- 27:48confidence i mean one of the
- 27:50and and the confidence comes by uh
- 27:53you know kind of by doing your homework
- 27:55right if if you go
- 27:57in thinking uh if you go into an
- 27:59interview thinking
- 28:00you are crazy if you don't hire me
- 28:04like you know you are insane i am the
- 28:06perfect person for this job
- 28:08because i've done this and this and this
- 28:09and this and i really
- 28:11think that um you know i can make a big
- 28:12impact here that is much more powerful
- 28:15than
- 28:16well i don't know if i'm gonna fit here
- 28:18and i'm not sure and you know
- 28:20so i think you know do the homework that
- 28:22you feel like
- 28:23you need to do to be prepared and and be
- 28:26over prepared when you're over prepared
- 28:28then
- 28:29then it shows it shows all the way
- 28:30through and if there's one thing that
- 28:32i've noticed about
- 28:32about people that succeeded apple it's
- 28:35it's a certain attitude
- 28:38of um they're there to solve
- 28:41the problems that mother nature has
- 28:44given us to solve
- 28:46right in order to deliver the next the
- 28:49next
- 28:49feature the next level of performance um
- 28:52they're not there for
- 28:53some to to to have them you know to wait
- 28:56around for someone to tell them what to
- 28:57do
- 28:58so a lot of curiosity i think is really
- 29:00important too
- 29:02there's no magic formula i mean i think
- 29:04there are a lot of venues
- 29:05for for this you know for this class
- 29:08there's the
- 29:09special pipeline there's the flash chat
- 29:11signups there's the
- 29:13applying throughout there's lots of this
- 29:14mechanics don't get too
- 29:16you know go for all the mechanics you
- 29:18know double the eyes cross all the t's
- 29:20but do your homework you know know your
- 29:23stuff and
- 29:24and get passionate about the technology
- 29:26and and the rules will come to you
- 29:28you know they'll be there when you when
- 29:30you pursue them
- 29:32and you know the truth the truth is and
- 29:34the truth is
- 29:35that for a lot of complicated hardware
- 29:40a really really useful thing for a
- 29:42designer to have is for lack of a better
- 29:45word a sense of style
- 29:47right that that you can you can
- 29:50you look at some pieces of of hardware
- 29:54and you say programming this is just
- 29:56ugly
- 29:57right this is a this and other ones you
- 29:59look at it and say
- 30:00the person who who specked this hardware
- 30:04understood how the software was going to
- 30:07feed it because it was clear they had
- 30:08written a lot of software
- 30:11to drive low-level hardware so don't
- 30:12underestimate the value of being a
- 30:14skilled embedded programmer
- 30:17a really skilled embedded programmer
- 30:20who's grounded in sort of the
- 30:21fundamentals
- 30:22of of the hardware underneath it is
- 30:24actually a pretty attractive candidate
- 30:27for some some hardware hardware tasks
- 30:30because you don't have to teach them
- 30:31style
- 30:35that that's a good one um here is a
- 30:37question um
- 30:38yeah are you looking for operative ee
- 30:40courses
- 30:43i was just typing an answer to them
- 30:47that's okay we can answer yeah i mean
- 30:51absolutely you know 140 151 152 those
- 30:55are great classes but you know
- 30:57uh but like like i was saying before i
- 31:00mean it's there's no like uh
- 31:02formula it's not like you're gonna you
- 31:03know go tick tick tick tick and
- 31:05and then you know your transcript has
- 31:07these things and then all of a sudden
- 31:08you know the the the you know the ufo
- 31:11lands in your backyard and the
- 31:12and the you know it comes down and and
- 31:15you know shovels your way to apple
- 31:16it doesn't work like that i mean a big
- 31:18part of what
- 31:20happens in those classes is you get to
- 31:22know people like borah better
- 31:24you know you get to know people like dan
- 31:26better or like her stay better or like
- 31:27ricky
- 31:28better you know go you know learn the
- 31:31craft in a deep
- 31:32way and start working with the people
- 31:33that are that are at berkeley that are
- 31:35so
- 31:36fantastic i mean you got you got you
- 31:38know ricky miller
- 31:39uh teaching analog circuits class with
- 31:4215 people on it that's criminal
- 31:44i would i would love to take that class
- 31:46with her with her i mean
- 31:48if you if i was your age whoever you are
- 31:51out there in zoom land if i was your age
- 31:53i would go take that class
- 31:55right now you know and whatever prereqs
- 31:57you know
- 31:58prereqs are guidance i would i would go
- 32:02to professors and say i what do i need
- 32:04to
- 32:06see now borah's never going to invite
- 32:07you back i you know i would go to
- 32:09professors and say i am so excited about
- 32:11this stuff
- 32:12what do i need to do to get to what i
- 32:14need to do to talk you into letting me
- 32:16take your class
- 32:17because if you're excited about it it
- 32:19comes through in every single way and it
- 32:20comes through in your style too just
- 32:21like dave was saying
- 32:23it comes through your style i can look
- 32:24at someone's schematic and tell if they
- 32:26were
- 32:26you know bored doing it or whether they
- 32:28cared about their craft you know
- 32:30it there's there is there's a key to
- 32:33this and it
- 32:34requires some degree of passion and i'm
- 32:36gonna shut up
- 32:38all right good thank you um here is
- 32:42another one
- 32:43what do you think have been the biggest
- 32:44successes and failures for apple's
- 32:47hardware group and what do you see as
- 32:50the main challenges going forward
- 32:53and feel free to also share about maybe
- 32:54the market wasn't ready
- 32:56like the newton was amazing but we
- 32:57weren't you know we weren't ready yet
- 32:59the market wasn't ready for it even
- 33:00though it was a great product feel free
- 33:01to share you know either marketing or
- 33:02engineering-wise in terms of that
- 33:04successes or values
- 33:10well i mean it re you know in in in
- 33:13recent times we've had we've we've had a
- 33:15pretty
- 33:16pretty good string of successes for sure
- 33:20um some of them beyond our wildest
- 33:22dreams
- 33:25we we've had things that internally have
- 33:27not worked out quite as
- 33:28quite as we would expect it and we don't
- 33:30talk about them very much
- 33:33not not because we're ashamed of them
- 33:35because sometimes
- 33:36they end up turning into something that
- 33:38actually works out really really well
- 33:41a little bit later um you know
- 33:45it's it's uh i i
- 33:48i i stopped
- 33:52counting the number of times we tried to
- 33:53build the ipad
- 33:56right the ipad the ipad that you know as
- 33:58the ipad
- 33:59um is not the first one it's not the
- 34:02second one it's not the third one
- 34:04um we would we built countless internal
- 34:07prototypes and
- 34:08and either it was too big or it was too
- 34:10heavy or the display didn't work or
- 34:13or using a pad with a stylus is a stupid
- 34:15idea
- 34:16or i mean you know that that and you'd
- 34:19build it and you'd say well that's
- 34:20interesting we learn some stuff but we
- 34:22but we just we don't declare it a
- 34:23failure we just put it on the shelf
- 34:26um until until something happens and
- 34:29then someone says well that's
- 34:30interesting
- 34:30let's take it off the shelf so and and
- 34:33frequently those failure internal
- 34:35failures turn into
- 34:36turn into successes sometimes you can
- 34:38learn as much from failures as you can
- 34:39from successes so you're like
- 34:41you actually usually learn you actually
- 34:44usually learn
- 34:45more right right from success
- 34:51when i do get to interview somebody i
- 34:53always like to talk about stuff that
- 34:55didn't work the discussion is usually
- 34:57um far more interesting and you learn a
- 34:59lot more about the person
- 35:01that's great that's right yeah and and
- 35:03you know there was a berkeley
- 35:05project that went from 1992 to 1995 that
- 35:08was called
- 35:09info pad
- 35:12it was a great success although not the
- 35:14commercial uh there was no immediate
- 35:16commercial transition i mean there are a
- 35:18few people
- 35:18you know the the guy that kind of was a
- 35:20driving student behind that is not a
- 35:22dean of engineering at mit so you can
- 35:24call that a success uh ananta
- 35:28uh chandragasan
- 35:31when i was a digital in the research
- 35:32labs uh hp
- 35:35announced that tiny little disk drive i
- 35:37think it was hiv
- 35:38yeah yeah hummingbird right yeah and we
- 35:41looked at that and said i know what you
- 35:42can do with this
- 35:43you can make a music player and we built
- 35:46this thing
- 35:47that had the disc and a signal processor
- 35:50and a bunch of mp3 decoding code that we
- 35:52licensed from frauhofer
- 35:54and a bunch of software for it and built
- 35:56this thing that it was about like eight
- 35:58inches by four inches by an inch and a
- 36:01half and it could store like 500 songs
- 36:03and it had shuffles and equalizers and
- 36:05all of those things
- 36:06and uh and it went nowhere
- 36:12actually actually that's not quite true
- 36:15um
- 36:15compact didn't know what to make of it
- 36:17and this korean company came to them and
- 36:19said
- 36:20we would like to license that and they
- 36:22said sure
- 36:23right yeah and compact the korean
- 36:26company licensed it
- 36:27and built like a hundred of them or
- 36:29something like this and tried to sell
- 36:30them
- 36:31and uh and nobody bought them
- 36:34and and it all seemed like like a like a
- 36:37total failure
- 36:37jared's laughing because he knows where
- 36:39this ends right um
- 36:41and then we built the ipod and the ipod
- 36:44was a great commercial success and when
- 36:46the ipod was a great commercial success
- 36:47all of these people would come come out
- 36:49of the woodwork and try to explain why
- 36:51they had this patent that
- 36:53that that read on the ipad and the
- 36:55lawyers would point to this little
- 36:56korean company and say
- 36:59that concept was previously available in
- 37:01commerce and
- 37:02and uh and and another patent would die
- 37:09all right it was really funny so i
- 37:12contributed to the iphone before i even
- 37:14worked for app
- 37:16before i worked for apple all right um
- 37:19a couple of very quick ones uh um i
- 37:22think this is for uh for a dave can you
- 37:24describe a little bit better
- 37:26what is good style
- 37:31um the short answer is no um
- 37:34you know you know when you see it right
- 37:36you know what you see
- 37:38you know it when you see it right you
- 37:40sort of look at it
- 37:41and you try to write some software to
- 37:43drive it and you realize that like
- 37:45like you know this device does not have
- 37:49some stupid read modify right bit
- 37:51right in the middle of a field that you
- 37:52have to change every time you start i o
- 37:55that require you know
- 37:56or something like that it's really hard
- 37:58to describe but you can sort of
- 37:59you can sort of tell it when you can
- 38:01when you see it right you see
- 38:03this is really really easy to talk to
- 38:05and the other thing is like
- 38:07everything is clumsy or the way the
- 38:09interrupt works is really crazy and you
- 38:11have to read these four registers and
- 38:13dismissing it's really hard
- 38:14and this as opposed to it just does the
- 38:17right thing
- 38:18right good style or not good style
- 38:21probably rears its head
- 38:23way way down the line when you run into
- 38:25scalability issues or you run into
- 38:27maintenance issues or
- 38:29all of a sudden you have to grow the
- 38:30team 10x in order to support the thing
- 38:32um good style good architecture those
- 38:35tend to go hand in hand
- 38:36and by the way there's a software
- 38:38equivalent which is just a good api i'm
- 38:39thinking about you're going to write
- 38:40some for some function some method some
- 38:42class
- 38:43what's the api that you create for it
- 38:45and style comes into that as well
- 38:47certainly
- 38:47oh style's all over the place in
- 38:49software i mean you can't you kidding me
- 38:51what do you when do you start
- 38:52documenting is that is that 20 minutes
- 38:54before you turn in the assignment
- 38:55you started your documentation or are
- 38:57you doing it all along
- 38:59you know i think i think the thing about
- 39:02style is that it shows
- 39:04uh and i saw the question here do you
- 39:06may try and make the hardware as neat
- 39:07and pretty as possible
- 39:09you know i i think the um the funny
- 39:11thing about that
- 39:12is it's it's symptomatic of the mindset
- 39:16of the person that created it
- 39:18how important is it to be
- 39:21for things to be done in a clean way
- 39:25right that means that i used to have a
- 39:27an instructor that said
- 39:29i am so lazy i will i will try a
- 39:32thousand times to make a proof that is
- 39:34one line shorter
- 39:35you know so so so but that
- 39:38is but the the idea there is to make a
- 39:40really clean
- 39:41implementation and to understand the
- 39:43value that there is
- 39:45not just in you know superfluous
- 39:47elegance but in terms of
- 39:48the quality of the work that's done so
- 39:50when you look at a
- 39:52when you look at a wire wrap board where
- 39:54you look on even a proto board
- 39:56and the the wires aren't all helter
- 39:58skelter and they're not like you know
- 40:00hanging on and there's you know you know
- 40:02you know by like one thread or something
- 40:04like that it means that there's
- 40:05attention put to detail
- 40:07um now if someone you know had it all
- 40:09held their sculpture and then they
- 40:11they shoved it all to make it look
- 40:12pretty because they knew the professor
- 40:14was running over i mean sure you can
- 40:15cheat on stuff like that
- 40:17that'll give you a little bit of a waste
- 40:18but but the but really
- 40:20the the more important thing is to get
- 40:22into the mindset of spending the amount
- 40:24of time
- 40:25to be careful and and cognizant and
- 40:28mindful of the different
- 40:29thing choices that you're making because
- 40:31you know especially in in programming
- 40:33oh my god the data structure is the most
- 40:35important thing you should spend ninety
- 40:37percent of your time on the data
- 40:38structure
- 40:40and then ten percent of your time on
- 40:41your code sit on your hands
- 40:44either that or you can write it five
- 40:46times you can write it the first time
- 40:47figure out that's wrong throw it all out
- 40:49write it a second time figure out that's
- 40:50a little bit better but still wrong
- 40:52throw that all out write it a third time
- 40:54you know and then
- 40:55and then eventually you're getting the
- 40:56hang of it right you know you're like
- 40:57okay well then
- 40:58that's what what you know you don't get
- 41:00to become an expert with that ten
- 41:01thousand hours right and you don't spend
- 41:03all the ten thousand hours
- 41:04you know making perfect things you make
- 41:06a lot of pretty crappy things along the
- 41:08way
- 41:12yeah it was a guy at deck i forget who
- 41:14it was
- 41:15but he used to always say that when he
- 41:18worked away on something for what seemed
- 41:20like an insufferable long time
- 41:22and when it was all done people would
- 41:23come in and look at the result and say
- 41:26how did that possibly take you that long
- 41:28to do this this is really easy
- 41:30right he knew that he had actually done
- 41:32a good job
- 41:36that might have been dave rogers who ran
- 41:38pdp 11 engineering for many years
- 41:42wow we mentioned some of these machines
- 41:44by doing the class
- 41:47um well um thank you very much uh thank
- 41:50you for coming stopping by
- 41:52um and this was very informative thank
- 41:54you dave
- 41:55guys we we had uh dave conrad we had
- 41:58jared zerby and we had
- 41:59anand uh of anantek himself in flesh
- 42:05so thank you again uh and we'll see more
- 42:07of you and we're gonna be distributing
- 42:09these prizes
- 42:10starting for project three right
- 42:13next week and um at least they're going
- 42:16to be
- 42:16paper certificates and then after we're
- 42:19done with project 4
- 42:20we're going to distribute those as well
- 42:23and
- 42:24um internships uh you guys are ready
- 42:27um these guys are telling you already
- 42:31and they want you so make sure to
- 42:32apologize make sure to apply definitely
- 42:35and have good things to say all right
- 42:37thank you folks that was wonderful nice
- 42:39to see you all
- 42:39thank you for your time all right
- 42:42take care
- 42:46all right all right good stuff
- 42:50uh we're up to the schedule um i wanted
- 42:53to actually
- 42:54do uh the schedule i think i moved to
- 42:56here here we go boom there i want to do
- 42:58computing the news
- 42:59so i have this delightful computer news
- 43:01which was uh it's just
- 43:02you know with all the insanity of the
- 43:04election it's nice to have something
- 43:05just fun and frivolous for the computing
- 43:07the news
- 43:08not like oh somebody hacked you know a
- 43:10voting system and oh god what does that
- 43:11mean for everybody
- 43:12thankfully the only news that i can i
- 43:14could find was this really delightful
- 43:15whimsical story
- 43:16which was this so they have now camera
- 43:19by the way if you have a if you have
- 43:20have a goal to be a career and as a
- 43:23camera operator i might
- 43:24encourage you to rethink your goals
- 43:26because ai is taking that job
- 43:28uh ai is also taking the job of
- 43:30radiologists radiologists job is to you
- 43:32know read a
- 43:33black and white picture and find out
- 43:35where the fracture is that's something
- 43:36that
- 43:36can be trained very easily um so watch
- 43:39out because he's a radiologist and he
- 43:41claims he is
- 43:42he is not going away well he is using ai
- 43:45he is using ai no no i appreciate i
- 43:47appreciate it i appreciate it but but i
- 43:49think the first pass maybe but being you
- 43:50know being done by
- 43:51by uh computers pretty soon so in this
- 43:54case in the camera operation job it's
- 43:56pretty easy all you do is you know
- 43:58you're
- 43:58there's a movement up and down do
- 43:59something like driving right i'm just
- 44:00doing this
- 44:01that's certainly something that a
- 44:02computer can do so you know it's moving
- 44:04left and right it's zooming
- 44:05and so they figured out how to write
- 44:06some software to control a camera and
- 44:08they had this it was doing well for
- 44:09for recording um so recording uh soccer
- 44:12matches
- 44:13and all you're trying to do is put the
- 44:14soccer ball in the center of the screen
- 44:16well it turns out there
- 44:17let me show you the picture this is a
- 44:19picture of what happened
- 44:21and um you see this is the this is a
- 44:23referee
- 44:24and the camera kept wanting to put the
- 44:26referee's head in the center of the
- 44:28screen
- 44:29the referee had a bald head and where he
- 44:31was standing on the sidelines
- 44:33his head is actually in the field like
- 44:35you know from the projection from the
- 44:36camera to there
- 44:37and then so he thinks well the ball is
- 44:38there and so they kept the entire match
- 44:41kept going back from the game to the
- 44:42referee and then like why is the camera
- 44:44keep showing the referee what's the deal
- 44:46and it turns out that the light is the
- 44:49lighting is just that way
- 44:50i mean it's exactly reflection exactly
- 44:52as you know the sun's in the far side
- 44:54anyway i want to share that
- 44:55very humorous thing um they confirmed it
- 44:58they confirmed it was a trouble
- 45:00that the zoom thought that they're gonna
- 45:01keep zooming in uh and they think if
- 45:03there was fixed shortly after the game
- 45:05ended blah blah blah but it was just a
- 45:06lightweight funny thing that
- 45:07the larger point of this is that machine
- 45:10learning systems
- 45:11your computer train systems can be very
- 45:13brittle uh and this was an
- 45:15example of a brittle system we see this
- 45:17with you know
- 45:18um facial recognition other things like
- 45:20this and so this is the case of
- 45:21recognition just being a little brittle
- 45:23so just having some fun with that
- 45:24all right schedule this is what our
- 45:27schedule looks like
- 45:27if you can see this let me zoom in a
- 45:29little bit so you can see here um we're
- 45:31at week 12 i've shown you week 11 behind
- 45:33us in week 13 ahead of us
- 45:34um we've got no lecture on wednesday so
- 45:37take a break on wednesday folks for the
- 45:38whole campus
- 45:39um rest to be charged if you can and i
- 45:41think in two weeks after that we're
- 45:43going to have
- 45:43[Music]
- 45:44so so enjoy your veterans day and happy
- 45:46veterans day to everybody
- 45:48um next week up thank you next week as a
- 45:51monday today's lecture is on cindy and
- 45:53flynn taxonomy
- 45:55on friday we start a whole new module so
- 45:57i want to encourage everybody who's
- 45:58behind and wants to kind of restart
- 46:00to jump in again uh really on you start
- 46:02on monday really because monday is
- 46:03really the first part of the section
- 46:04this whole section on parallelism this
- 46:06is the topic you see the bottom part of
- 46:07my slide here
- 46:08it says parallelism this is the topic of
- 46:09parallelism so please do
- 46:11jump in if you were behind you can you
- 46:12can catch up here with 32 33
- 46:1534 35 and then the fun thing is we're
- 46:17going to be adding 36 and 37 which is
- 46:19mapreduce
- 46:20spark and warehouse scale computing so
- 46:23how can you
- 46:24you know literally with a couple of
- 46:26lines of software a couple lines of
- 46:27python
- 46:28control a million machines really
- 46:30amazing stuff that
- 46:31these abstractions are just remarkable
- 46:33so we'll share that
- 46:34uh the important comments about that uh
- 46:38project four depends basically you know
- 46:40you can start project for
- 46:42with your c knowledge you can get a good
- 46:44head start with
- 46:46a project yeah uh there was a good
- 46:49guideline um guidance through the
- 46:51navigation through the project four that
- 46:53was done by g
- 46:54i think it's recorded and please go
- 46:56ahead and take a look at that
- 46:58um you know in order to excel in it you
- 47:00basically need to start with a monday's
- 47:02lecture
- 47:02and then you know that is a little bit
- 47:05longer but i really wanted to squeeze in
- 47:07these things that are important for the
- 47:08project
- 47:10and then you have um i think about two
- 47:13lectures from the
- 47:14uh dan um uh fridays and mondays we'll
- 47:17get you everything you need to do
- 47:19uh for the project maybe wednesdays as
- 47:22well yeah maybe a little wednesday as
- 47:23well good stuff
- 47:24good stuff all right great so now let's
- 47:26take a look at
- 47:27where we are with our schedule again so
- 47:29this is another another
- 47:30pass on that so we're in um
- 47:33parallelism holiday parallelism tlp one
- 47:36two and three and then my produce and
- 47:37spark
- 47:38in lecture in terms of homework no this
- 47:39is new information this is all
- 47:41just copying and pasting from the
- 47:42website into this just so you know but
- 47:44seven and eight we moved seven and eight
- 47:45we pushed it down and slid it to today
- 47:47so they're due today hopefully that
- 47:48isn't too
- 47:49too too bad to be able to put them
- 47:50together i've been told that it doesn't
- 47:51take that long to finish with both of
- 47:52those
- 47:53and hopefully you've been doing that um
- 47:55we've removed set 9 and 11 although they
- 47:57still show on the website because they
- 47:58were going to be for practice only we're
- 47:59going to remove the dates
- 48:00we're going to make sure it's there but
- 48:01this is not new information just make
- 48:02sure you know
- 48:03that 9 and 11 are just for practice only
- 48:05so get some gets
- 48:06get some work there play with that you
- 48:08won't need them for the exams that was a
- 48:10question people were asking
- 48:11you will not need for exams all the
- 48:12material all the things you learned
- 48:14there
- 48:14we'll also be teaching you a lecture and
- 48:16the exam will be tested on material that
- 48:18you could have gotten from lecture so
- 48:19you don't need a homework in 911
- 48:21for the exam it's not like it's really a
- 48:23secret thing that's there
- 48:25that's different from labs again this is
- 48:26just a clarification of a previous
- 48:28policy
- 48:28lab the material is on is due on the
- 48:32exams
- 48:32but we're by not having you have to
- 48:34check to do a check-in
- 48:36we're basically giving you flexibility
- 48:37to be able to do those labs anytime you
- 48:38want to
- 48:39uh and ask our tas whenever you have any
- 48:41questions be able to ask them but it's
- 48:42out there it's like
- 48:42it's like reading the book so labs
- 48:44become reading the book that's the book
- 48:45is also on the exam so make sure you do
- 48:46that
- 48:47we also took a look at the project for
- 48:49deadline and we moved that deadline back
- 48:52maybe we'll paste this again into piazza
- 48:53people don't watch these videos
- 48:55but project four we moved them back so
- 48:56you can enjoy thanksgiving it was gonna
- 48:58be due
- 48:58the monday after thanksgiving and no
- 49:00one's gonna i mean that that's gonna be
- 49:01a painful thing for everybody in terms
- 49:03of thanksgiving so we slid it back
- 49:05to be due 12-2 and you have a little bit
- 49:07of time in the next video it's not i
- 49:08think that's due wednesday or thursday i
- 49:09think 12 2 is either wednesday or
- 49:10thursday
- 49:11wednesday it is wednesday we don't have
- 49:13a ton of slippage after that
- 49:15because you know there are some people
- 49:16who have uh extra time and you know
- 49:19we have six flip days yeah we couldn't
- 49:21we could go into you know
- 49:22until 2021 we've all the slip days so we
- 49:24did that um
- 49:26and this is also an existing information
- 49:28just we're just saying it again um the
- 49:30midterm retake was pushed to 11
- 49:3115 we have 25 11 15 is a sunday you've
- 49:34got 24 hours and we're going to be
- 49:36letting you know
- 49:37very early this week cc's working on the
- 49:39on the file we're going to try to make a
- 49:41pass on this
- 49:42and declare very early this week who
- 49:43needs to take it uh and then we'll let
- 49:45you have a chance to say
- 49:46well i wasn't on the list but i want to
- 49:47be on the list or i wasn't i was on the
- 49:49list but i don't think i should be on
- 49:50the list because you can't find my
- 49:51videos so we'll try to
- 49:52have another pass where we'll declare a
- 49:54list of students we think should take it
- 49:56then we'll give you a chance to either
- 49:57add you or remove yourself from the list
- 49:58we'll make a call on that and then give
- 49:59you the final list and that's what we're
- 50:00that's
- 50:01that's what we'll expect to take the
- 50:02exam but that's it and again we only
- 50:03have to take the exam for pieces that
- 50:04you missed
- 50:05uh not for the whole exam unless
- 50:07everything's messed up in terms of your
- 50:08video in which case you might ask you to
- 50:09hold the whole thing done
- 50:10but that's it all right that's it
- 50:13that's it any questions i'm looking at
- 50:15the q a i see nothing
- 50:17i'm looking at here yeah i think we're
- 50:20pretty clear
- 50:21there's no information so it's not like
- 50:22well we made some new policy and just
- 50:24questions about it so i think we're
- 50:25pretty
- 50:25pretty clean there but um and i i have a
- 50:28collection of some internship postings
- 50:29i'm gonna stop
- 50:30posting them on piazza and um you guys
- 50:33are ready for many of those i mean
- 50:34you're in a good shape
- 50:37go for it good stuff and thank you by
- 50:39the way for attending i see about about
- 50:4050
- 50:4155 or 56 people were here today so
- 50:43thanks folks for coming to this
- 50:44um let's see
- 50:49we we had a lot of slip days we were not
- 50:52done
- 50:52grading so no paper certificates today
- 50:55yet
- 50:55uh we are not done checking everything
- 50:58out
- 50:59from project 3b yep
- 51:02good stuff all right all right folks
- 51:06all right pretty well thanks everybody
- 51:08good to see you all
- 51:10we'll see you next week all right take
- 51:11care folks bye-bye thanks
- 51:13bye
About this transcript
This page contains the full transcript of [CS61C FA20] Weekly Lecture 12.LIVE - SIMD by CS 61C Departmental, generated from the public captions YouTube serves with the video. The transcript has 9,886 words across 1,590 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.