Please enable Javascript, which is required for this site.
async-parallel-app-design

Async and Parallel Programming: Application Design

Discusses the design of asynchronous and parallel applications using the new Task-based model available in .NET 4 and Silverlight 5.

Authored by: Dr. Joe Hummel
Duration: 3h 45m
Level: Intermediate
Released: 5/29/2012
Features: Closed captions
Course Rating:
You are currently not signed in. Please sign in to access subscriber-only content.
Currently using:  {{getCurrentPlayerName()}}  [Change]
expand all | collapse all Progress Duration
This module is unavailableUnderstanding the Dangers of Concurrency
 You have not watched this Module. 01:23:11
Introductions and Agenda
You have not watched this Clip. 01:52
The pitfalls of concurrency
You have not watched this Clip. 01:50
Correctness guarantees: safety, liveness
You have not watched this Clip. 00:44
Terminology: race conditions, critical sections
You have not watched this Clip. 00:54
Race conditions involving shared resources
You have not watched this Clip. 03:36
Demo 1: parallelizing a Logfile Search app
You have not watched this Clip. 01:14
Demo 1: source code layout
You have not watched this Clip. 04:26
Demo 1: from sequential to parallel
You have not watched this Clip. 11:41
Demo 1: summary of why it's broken
You have not watched this Clip. 00:18
Solving race conditions --- an overview
You have not watched this Clip. 01:00
Solution 1: locking
You have not watched this Clip. 03:57
Demo 2: solution using locking
You have not watched this Clip. 03:56
Solution 2: interlocking
You have not watched this Clip. 01:14
Demo 3: solution using interlocking
You have not watched this Clip. 07:53
Solution 3: lock-free
You have not watched this Clip. 02:14
Demo 4: lock-free solution
You have not watched this Clip. 13:35
The danger of shared objects
You have not watched this Clip. 01:55
Thread-safety and solutions to shared objects
You have not watched this Clip. 02:54
Demo 5: using thread-safe objects
You have not watched this Clip. 04:14
Synchronization primitives
You have not watched this Clip. 01:07
The concurrent data structures: thread-safe collections
You have not watched this Clip. 00:22
Performance lessons...
You have not watched this Clip. 02:21
Demo 6: improving performance
You have not watched this Clip. 08:31
Summary and References
You have not watched this Clip. 01:23
This module is unavailableExecution Model and Types of Parallelism
 You have not watched this Module. 01:04:19
Introductions and Agenda
You have not watched this Clip. 01:59
Tasks: developer's role vs. .NEt's role
You have not watched this Clip. 00:24
Execution model: high-level view
You have not watched this Clip. 01:29
Task granularity: just how lightweight?
You have not watched this Clip. 00:58
Execution model: tasks vs. threads vs. cores
You have not watched this Clip. 02:33
Custom task scheduling
You have not watched this Clip. 00:53
Additional observations regarding task scheduler
You have not watched this Clip. 01:35
Work-stealing in detail
You have not watched this Clip. 01:52
Two important assumptions: short-lived, random order
You have not watched this Clip. 00:50
Fairness option to execute tasks in order
You have not watched this Clip. 01:22
LongRunning option for tasks > 1-2 seconds
You have not watched this Clip. 00:45
Demo 1: executing 100 long-running tasks
You have not watched this Clip. 00:30
Demo 1: solution attempt with default behavior
You have not watched this Clip. 04:26
Demo 1: solution attempt with long-running option
You have not watched this Clip. 02:08
Solution for 100 long-running tasks
You have not watched this Clip. 01:47
Demo 2: proper solution in action
You have not watched this Clip. 04:32
Parallel.For = more concise solution
You have not watched this Clip. 00:44
Types of Parallelism
You have not watched this Clip. 00:36
Data parallelism
You have not watched this Clip. 00:52
Task parallelism
You have not watched this Clip. 00:37
Dataflow parallelism
You have not watched this Clip. 02:24
Embarrassingly parallel
You have not watched this Clip. 00:57
TPL support for parallelism types
You have not watched this Clip. 00:25
Parallel.For, .Foreach, .Invoke
You have not watched this Clip. 02:44
Structured (fork-join) parallelism
You have not watched this Clip. 01:07
Demo 3: Mandelbrot app overview
You have not watched this Clip. 01:11
Demo 3: sequential version
You have not watched this Clip. 01:57
Demo 3: Parallel.For version
You have not watched this Clip. 06:38
Data partitioning
You have not watched this Clip. 02:15
Custom data partitioning
You have not watched this Clip. 02:28
Exception handling with Parallel class
You have not watched this Clip. 00:41
Breaking out of a Parallel loop
You have not watched this Clip. 01:26
Cancelling a Parallel loop
You have not watched this Clip. 01:40
Demo 4: Mandelbrot app with cancellation
You have not watched this Clip. 05:50
Summary and References
You have not watched this Clip. 01:44
This module is unavailableDesigns and Patterns for Parallel Programming
 You have not watched this Module. 01:17:32
Introductions and Agenda
You have not watched this Clip. 01:35
Design Problem 1: 100+ CPU-intensive operations
You have not watched this Clip. 02:25
Design Problem 2: download 20+ web pages
You have not watched this Clip. 03:14
Design Problem 3: application logging task
You have not watched this Clip. 02:09
Parallel Patterns --- master list
You have not watched this Clip. 00:28
Pipeline pattern
You have not watched this Clip. 01:29
Dataflow pattern
You have not watched this Clip. 01:34
Increasing parallelism in pipeline and dataflow
You have not watched this Clip. 01:16
Concurrent Data Structures --- master list
You have not watched this Clip. 00:59
ConcurrentQueue T
You have not watched this Clip. 03:31
Demo 1: overview of Netflix data mining app
You have not watched this Clip. 03:24
Demo 1: overview of parallel solution (v1)
You have not watched this Clip. 02:44
Demo 1: implementation with ConcurrentDictionary T
You have not watched this Clip. 05:43
Demo 1: results
You have not watched this Clip. 01:14
Producer-Consumer pattern
You have not watched this Clip. 01:23
Implementation of Producer-Consumer with BlockingCollection T
You have not watched this Clip. 03:03
Demo 2: Netflix solution using Producer-Consumer (v2)
You have not watched this Clip. 07:40
Demo 2: results
You have not watched this Clip. 00:52
MapReduce pattern
You have not watched this Clip. 00:51
Implementing MapReduce
You have not watched this Clip. 04:24
Demo 3: Netflix using MapReduce
You have not watched this Clip. 03:50
Demo 3: results
You have not watched this Clip. 01:33
Parallel LINQ (PLINQ)
You have not watched this Clip. 02:46
Demo 4: Netflix with PLINQ
You have not watched this Clip. 03:08
Speculative Execution pattern
You have not watched this Clip. 01:25
APM: Asynchronous Programming Model pattern
You have not watched this Clip. 01:23
APM example of async file I/O
You have not watched this Clip. 01:17
Demo 5: Stock History with Speculative Execution and APM
You have not watched this Clip. 09:01
Parallel I/O design challenge
You have not watched this Clip. 01:45
Summary and References
You have not watched this Clip. 01:26