Threading Building Blocks from Intel

Intel® Threading Building Blocks (Intel® TBB) is a widely used, award-winning C++ template library for creating reliable, portable, and scalable parallel applications. Use Intel® TBB for a simple and rapid way of developing robust task-based parallel applications that scale to available processor cores, are compatible with multiple environments, and are easier to maintain. Intel® TBB is the most proficient way to implement future-proof parallel applications that tap into the power and performance of multicore and manycore hardware platforms.

Available as open source

• Enhanced Productivity and Reliability - Intel® TBB provides abstractions that make it easier to get scalable and reliable parallel applications with fewer lines of code

• Scalability with Future-proofing - Application performance automatically improves as processor core count increases by using abstract tasks

• Comprehensive - Intel TBB provides a set of components needed to implement parallel code

• Portability - Intel TBB runs on a wide variety of OSs and platforms

Intel® Threading Building Blocks is also available within development tool suites, including:

• Intel® Parallel Studio XE

• Intel® Composer XE

• Intel® Cluster Studio XE

Intel Threading Building Blocks uses task-based abstractions that make it easier to get scalable and reliable parallel applications with fewer lines of code. Task-based algorithms, containers, and synchronization primitives simplify parallel application development.

Side-by-side comparison on the left shows equivalent Windows* thread functionality that requires significantly more code to make a 2D ray tracing program, Tachyon, correctly threaded. Linux* and Mac OS X* developers can expect similar results.

Application performance automatically improves as processor core count increases by using abstract tasks. Sophisticated task scheduler dynamically maps tasks to threads to balance the load among available cores, preserve cache locality, and maximize parallel performance.

Intel® Threading Building Blocks yields linear scaling on a 40 core system in these three example applications

Intel Threading Building Blocks provides a suite of components that help simplify and speed development of scalable, high-performance parallel applications. It provides parallel algorithms, concurrent containers, task scheduling, scalable memory allocation, and synchronization primitives. Version 4.0 expands usage models and improves performance and usability. Text in bold italics indicates new functions for v4.0.

Intel Threading Building Blocks is validated and commercially supported on Windows*, Linux*, and Mac OS* platforms. It is also available on FreeBSD*, IA Solaris*, XBox* 360, and PowerPC-based systems via the open source community. Organizations can expand their customer base by using a production-ready, open solution for parallelism that is available on a broad range of platforms.

September 8, 2011: Announcing expanded usage models and improved performance and usability with TBB 4.0

December 23, 2010: Announcing the Graph Community Preview (CP) Feature

Following on the recent introduction of the first CP feature, the concurrent_priority_queue, the Intel® Threading Building Blocks (Intel®TBB) team introduced a new feature today in Intel TBB 3.0 Update 5 -- the Graph interface. The Graph interface is the culmination of years of work to meet the needs of developers who would like to utilize Intel TBB for irregular problem types like those that use events or message passing. We are particularly interested in community feedback on whether the graph is helping produce more elegant and high-performing implementations in real-world applications. As always, community questions, feedback, and contributions are welcome.

Useful links:

• Read about the Graph interface and its many uses in Mike Voss's blog series: Using the Intel® Threading Building Blocks Graph Community Preview Feature

• Download Intel TBB 3.0 Update 5

December 8, 2010: Announcing Community Preview (CP) Features

The Intel® Threading Building Blocks (Intel TBB) team is happy to introduce the use of Community Preview (CP) features into Intel TBB today. CP features are a great way for Intel to show new and interesting capabilities to our community and customers before they have been finalized. As part of our commitment to openness with all of the Intel® Parallel Building Blocks technologies, we want our users to know what we are working on to make Intel TBB better. We also want to gain your feedback on upcoming features so that we can make sure we continue to meet your needs today and in the future.

These features are fully tested but are not officially supported or necessarily fully documented. Given the early nature of these features, we don't guarantee that they won't be removed or modified in ways that break compatibility with pre-production versions. In addition, they are turned off by default so they won't impact your application unless you want to give them a try. We look forward to hearing your response in the forum to our first CP feature, the Concurrent Priority Queue, and to the use of CP features in general.

Useful links:

• Read about the Community Preview features in Victoria Gromova's knowledge base article: Intel® Threading Building Blocks: Community Preview Features

• Read about the Concurrent Priority Queue in Terry Wilmarth's blog: Intel® Threading Building Blocks Version 3.0 Update 4 showcases its first Community Preview feature: Concurrent Priority Queue

• Download Intel® Threading Building Blocks 3.0 Update 4

Enhancements in Intel Threading Building Blocks 3.0

• Extended compatibility and interoperability support for Microsoft Visual Studio* 2010 Parallel Patterns Library* (PPL) and Concurrency Runtime* (ConcRT)
• Added Microsoft Windows* 7, and Apple Mac OS* Snow Leopard support
• Enhanced Task Scheduler provides starvation-proof scheduling tasks for queue-like work; Master Thread Isolation improves task-scheduling predictability and responsiveness
• Enhanced Memory Allocator includes performance optimization for large-block allocations
• Expanded lambdas support provides C++ 0x condition variable; parallel_pipeline a strongly typed lambda friendly interface for building and running pipelines; new container concurrent_unordered_map support