Apple's Grand Central Dispatch: Path to Multicore
Apple has taken multicore to heart with Grand Central Dispatch, a deeply ingrained capability of their Snow Leopard OS to really help with using multicore parallelism. I'm a big fan, including the fact that it is not all new.
There are many sessions at WWDC (Apple's developer conference) which promise to be quite informative. Apple is really helping unlock the power of multicore processors.
What Is It?
At the heart of Grand Central Dispatch is NSOperationQueue which debuted in "Leopard" (released October 2007). Around this simple "hey, throw this work on the queue" is a commitment in the operating systems and tools to encourage its use and make it easy to access.
On Friday of this week, after attending sessions this week from Apple on Grand Central Dispatch at WWDC, I recommend not missing a session which show Grand Central Dispatch in action and comparing it with the same program using TBB.
WWDC Session 131: Scaling Performance Using Grand Central Dispatch and Intel Libraries, Friday 2:00-3:15 pm
Session 131 Abstract: Scaling the performance of your application to get the most out of a multicore Mac requires an understanding of the various tools and libraries available at your disposal. Intel engineers will demonstrate, with the help of a real-world use case, how you can employ Apple's Grand Central Dispatch, Intel's Threading Building Blocks and Intel's Performance Primitives to achieve and scalable application performance on Mac OS X.Intel Presenters: Pallavi Mehrotra, Richard Hubbard
Pallavi and Richard have a very education talk that will make you wanting to learn even more!
Will GCD replace TBB?
No, because that's not what it's about. In fact, it may be worth porting TBB on top of GCD one day. We'll see what the TBB open source project decides. GCD is an enabling technology for applications and tools. An application can use GCD directly with great affect. Pallavi and Richard show that in their talk.A Simple Answer?
I'd use GCD when programming in Objective-C, or TBB when programming in C++ or C. Session 131 shows that either choice can work well on Mac OS X.A Little Deeper Answer?
TBB has more knobs, is cross-OS portable, and addresses issue specific to C++.
GCD has deep support from Apple in tools and Mac OS X, and has been well integrated into Objective-C. I'm sure this will make a fertile ground for more blogging in the future. This week, it's time to learn at WWDC.
Parallel Pattern 5: Stencil
All memory addresses used for reads are expressed as offsets
Distributing Work Across Cores Using .NET
A roll-your-own ThreadPool implementationLooking For The Lost Packets: Part 2
Looking For The Lost Packets: Part 1
- Intel Parallel Studio; Download the free eval today!
- Parallelism Breakthrough Video Series; Watch and learn more about Intel® Parallel Studio
- 2009 Intel Software Webinar Series; View On-Demand webinars
- Coding for Multi-core Processes; Intel® Compiler Pro eBook
- Performance Through Parallelism; Intel® Tuning for Vista eBook
- Intel® Software Network; Connect with developers and Intel engineers
-
February 18, 2010
Lock Contention, Using Intel Parallel Studio to Improve Performance
Speaker: Vasanth Tovinkere, Software Engineer, Intel Corporation (Bio)Vasanth Tovinkere is a software engineer in the Developer Products Division (DPD) at Intel. His current role involves defining novel approaches to understanding and visualizing parallel performance and consulting with strategic customers to help them prepare and deliver code for the multicore world. Vasanth has been involved in the development of automatic semantic event detectors for digital sports technologies in Intel Labs. He also has been awarded three patents and has two patents pending.
Abstract:
Discover how easy it is to use the power of Microsoft Visual Studio and Intel Parallel Studio to find performance issues due to lock contention in threaded applications. This ensures that shipped applications can take better advantage of multicore processors. In this webcast, we provide live demonstrations that show how to identify lock contentions issues with Visual Studio and Intel Parallel Studio, an add-in to Visual Studio that helps developers create fast, reliable code on multicore processors.t.



