Weiter zum Menü

Fachgruppe Kastens > Lehre > Parallel Programming WS 2014/2015 > Slides

Weiter zum Inhalt

• Home
• First Slide
• Last Slide
• Chapters
• Print Slides
• Slides (plain text)

SUCHEN:

Weiter zum Ende

Parallel Programming WS 2014/2015 - Slides

0. Introduction [-]
Parallel Programming (01) Objectives (02) Contents (03) Prerequisites (04) Organization of the course (05) Literature (06) Parallel processes (07) States and transitions of processes (08) Applications of parallel processes (09) Create threads in Java - technique: implement runnable (10) Create threads in Java - technique: subclass of Thread (11) Important methods of the class Thread (12) Example: Digital clock as a process in an applet (1) (13) Example: Digital clock as a process in an applet (2) (14)
1. Properties of parallel programs [-]
2. Properties of Parallel Programs (15a) Proofs of parallel programs (15ab) Hoare Logic: a brief reminder (15b) Axioms and inference rules for sequential constructs (15c) Weakest precondition (15d) Weakest precondition (15e) Examples for weakest preconditions (15f) Interleaving - used as an abstract execution model (17a) Atomic actions (17b) Atomic by AMO (17c) Interference between processes (17d) Example: Interference between an assertion and an assignment (17e) Non-interference checks (17f) Two inference rules for concurrent execution (17g) Avoiding interference (17h)
2. Monitors in general and in Java [-]
Communication and synchronization of parallel processes (18) Monitor - general concept (19a) Condition variables (19b) Example: bounded buffer (20) Synchronization in a monitor (21) Variants of signal-wait semantics (22) Variants of signal-wait semantics: examples of execution (22a) Monitors in Java: mutual exclusion (23) Monitors in Java: condition synchronization (24) A Monitor class for bounded buffers (25) Concurrency Utilities in Java 2 (25j) Concurrency Utilities in Java 2 (example) (25k)
3. Systematic development of monitors [-]
Design of monitors: monitor invariant (26) Systematic design of monitors (27) Derivation step 1 (27a) Derivation step 2 (27b) Derivation step 3 (27c) Derivation step 3 (27ca) Derivation step 4 (27d) Derivation step 4 (27da) Derivation step 5 (27e) Relevant state changes (28) Pattern: Allocating counted resources (29) Monitor for resource allocation (30) Processes and main program for resource monitor (31) Readers-Writers problem (Step1)(32a) Readers-Writers problem (Step2) (32b) Readers-Writers problem (Step3) (32c) Readers-Writers problem (Step4) (32d) Readers-Writers problem (Step5) (32e) Readers/writers monitor (33) Method: rendezvous of processes (34) Monitor design for the Sleeping Barber problem (step 1)(35) Monitor invariant for the Sleeping Barber problem (step 2)(35) Waiting conditions for the Sleeping Barber problem (step 3)(36) Substitute counters (step 3a)(37) Signal operations for the Sleeping Barber problem (step 4)(37a)
4. Data parallelism: barriers [-]
5. Data parallelism: Barriers (38) Data parallelism as an architectural model (39) Data parallelism as programming model (40) Example prefix sums (41) Example prefix sums (2) (41a) Prefix sums: applied methods (42) Barriers (43) Barrier implemented by a monitor (44) Distributed tree barrier (45) 2 Rules for Synchronization Using Flags (45a) Distributed tree barrier implementation (45b) Symmetric, distributed barrier (dissemination) (46) Symmetric, distributed barrier: implementation (46a) Prefix sums with barriers (47) Prefix sums in a synchronous parallel programming model (48) Find list ends data parallel (49)
5. Data parallelism: loop parallelization [-]
6. Data Parallelism: Loop Parallelization (50) Iteration space of loop nests (51) Examples for Iteration spaces of loop nests (51a) Data dependences (52) Loop transformation (53) Transformation of data and loop nests (54) Transformations defined by matrices (55) Reversal (55a) Skewing (55b) Permutation (55c) Using transformation matrices (56) Inequalities Describe Loop Bounds (56a) Transformation of Loop Bounds (56b) Example for Transformation and Parallelization of a Loop (56c) Solution of the Transformation and Parallelization Example (56d) Transformation and parallelization (57) Data distribution (58) Data distribution for parallel loops (59)
6. Asynchronous message passing [-]
Asynchronous message passing (60) Channels implemented in Java (61) Processes and channels (62) Termination conditions (62a) Message structures (63)
7. Client-Server paradigm [-]
Client / server: basic channel structure (64) Server processes: different kinds of operations (65) Different kinds of operations on different channels (66) Several servers (67) Receive without blocking (68) Conversation sequences between client and server (69) Active monitor (server) vs. passive monitor (70)
8. Messages in distributed systems [-]
Distributed processes: Broadcast in a net of processors (71) Broadcast method (72) Probe and echo in a net (73) Probe and echo: detailed operations (74) Connections via ports and sockets (75) Sockets and I/O-streams (76) Worker paradigm (77) Branch and Bound (78) B&B example: Travelling sales person (78a) Parallel Branch & Bound (central) (79) Parallel Branch & Bound (distributed) (80) Termination in a ring (81) Method calls for objects on remote machines (RMI) (82) RMI in Java (83) RMI development steps (84) RMI development steps (continued) (85) Objects as parameters of RMI calls (86)
9. Synchronous message passing [-]
Synchronous message passing (87) Notations for synchronous message passing (88) Selective wait (89) Guarded loops (90) Prefix sums computed with synchronous messages (91) No deadlocks in synchronous prefix sums (92) Client/Server scheme with synchronous messages (93) Synchronous Client/Server: variants and comparison (94) Synchronous messages in Occam (94a) Bounded Buffer in Occam (94aa) Synchronous rendezvous in Ada (94b) Ada: Synchronous rendezvous (94ba)
10. Concurrent and functional programming [-]
10. Concurrent and functional programming (94c) Recursive adaptive quadrature computation (94d) Recursive adaptive quadrature computation (94e) Streams in functional programming (94f) Examples for stream functions (1) (94g) Examples for stream functions (2) (94h) Recursive stream composition (94i) Sieve of Eratosthenes in CSP (94j) Dataflow languages (94k) Language Erlang (94l) Basic communication constructs (94m) Complete example: Counter (94n) Example: Allocation server (interface) (94o) Example: Allocation server (implementation) (94p) Scala: object-oriented and functional language (94q) Actors in Scala (1) (94r) Actors in Scala (2) (94s)
11. Questions [-]
11. Check your knowledge (1) (95) Check your knowledge (2) (95a) Check your knowledge (3) (96) Check your knowledge (4) (97) Check your knowledge (5) (98)

Generiert mit Camelot | Probleme mit Camelot? | Geändert am: 11.01.2015

Weiter zum Anfang