8. Analysis of Multithreaded Algorithms

  Рет қаралды 27,271

MIT OpenCourseWare

MIT OpenCourseWare

Күн бұрын

Пікірлер: 18
@chuanqisun
@chuanqisun 2 жыл бұрын
To me, the most valuable part of the lecture starts from 31:10. Charles' philosophy on "empathetic education" truly represents the value of MIT and is unfortunately missing from so many institutions. Thank you for this lesson
@effy1219
@effy1219 10 ай бұрын
I really luckily randomly clicked in that time slice, and have the best lesson ever
@leixun
@leixun 3 жыл бұрын
*My takeaways:* 1. Algorithm revision 0:40: divide-and-conquer recurrences 2. Cilk loop parallelism 15:31 3. Three performance tips 1:03:51 4. Matrix multiplication 1:07:09
@jaggis4914
@jaggis4914 3 жыл бұрын
Thanks Prof. Leiserson! Thank you MIT
@dankal444
@dankal444 6 ай бұрын
During studies I was called "Question Man" and not without a reason. Lecturers were not accustomed to students asking questions so I had various responses. Mostly positive, must say :) Only on one subject I was being treated as if I was asking silly, unnecessary questions. But truly, half the math being used was new to us or not used since 2 years ago in math courses. During exams it turned out I was among a few that passed. Waste of time and energy
@pschneider1968
@pschneider1968 2 жыл бұрын
This guy is a great teacher!
@gurpreetsinghbains5470
@gurpreetsinghbains5470 4 жыл бұрын
Love this man
@turun_ambartanen
@turun_ambartanen Жыл бұрын
I have a question about 1:00:40, puny parallelism. I don't understand why we don't get any parallelism. In my mind, Work is T_1 = n*I + n*S and Span is T_inf = n*S + I. Then the Parallelism has two limits, one where spawning is relatively expensive (S>>I), in which case it comes out to be T_1/T_inf = 1 as shown on the slide, but when spawning is relatively cheap (I>>S), it comes out to be T_1/T_inf = n. Obviously the example is the first case, resulting is Puny Parallelism, but say you parallelize sleep(1000) calls, it is quite obvious we do get a speedup. Am I wrong? Or is the result on the slide only for the example shown on the slide? Please correct me if I made any mistakes!
@anon-j6f
@anon-j6f Жыл бұрын
ive met you. im totally going to have a ton of questions for you and your colleagues next time im in town.
@hkkhgffh3613
@hkkhgffh3613 5 жыл бұрын
Booyakasha! Respect Prof. Leiserson! Skol!
@thereGoMapo
@thereGoMapo 4 жыл бұрын
ah university lectures, where hand wavy proofs abound
@u2bacount
@u2bacount 2 жыл бұрын
brilliant
@johnpapa8554
@johnpapa8554 2 жыл бұрын
how much does it cost for that genious to open my eyes?? does anyone know please?
@hkkhgffh3613
@hkkhgffh3613 5 жыл бұрын
Da whiteboard is below par. Better cleaning required!
@albertovalsania8656
@albertovalsania8656 5 жыл бұрын
Is possibile to get italiano subtitles?
@hectorcanizales5900
@hectorcanizales5900 3 жыл бұрын
Yes, the KZbin auto translator is probably your best choice and usually works fairly good
@dankal444
@dankal444 6 ай бұрын
During studies I was called "Question Man" and not without a reason. Lecturers were not accustomed to students asking questions so I had various responses. Mostly positive, must say :) Only on one subject I was being treated as if I was asking silly, unnecessary questions. I felt stupid. But truly, half the math being used was new to us or not used since 3 years ago in math courses. During exams it turned out I was among a few that passed. Waste of time and energy
9. What Compilers Can and Cannot Do
1:18:46
MIT OpenCourseWare
Рет қаралды 77 М.
10. Measurement and Timing
1:21:28
MIT OpenCourseWare
Рет қаралды 18 М.
Kluster Duo #настольныеигры #boardgames #игры #games #настолки #настольные_игры
00:47
Synyptas 4 | Жігіттер сынып қалды| 3 Bolim
19:27
kak budto
Рет қаралды 1,2 МЛН
Сюрприз для Златы на день рождения
00:10
Victoria Portfolio
Рет қаралды 1,5 МЛН
10. Understanding Program Efficiency, Part 1
51:26
MIT OpenCourseWare
Рет қаралды 236 М.
Java 21 new feature: Virtual Threads #RoadTo21
33:35
Java
Рет қаралды 65 М.
17. Synchronization Without Locks
1:20:10
MIT OpenCourseWare
Рет қаралды 33 М.
Faster than Rust and C++: the PERFECT hash table
33:52
strager
Рет қаралды 582 М.
4. Assembly Language & Computer Architecture
1:17:35
MIT OpenCourseWare
Рет қаралды 718 М.
The Tragedy of systemd
47:18
linux.conf.au
Рет қаралды 1,1 МЛН
Kluster Duo #настольныеигры #boardgames #игры #games #настолки #настольные_игры
00:47