When refactoring a serial application into a parallel application the minimum bar of acceptable performance is that the parallel application must run from beginning to end in less wall clock time than the serial application. If it doesn’t run faster, there’s no point in doing the refactoring...

As interesting as profiling applications on your development computer is, I’m sure you’ve wanted to see the behavior of applications when running in production systems. The ability to view remote traces with the Concurrency Visualizer first requires installing the Visual Studio 2010 profiler...

This is the picture that Concurrency Visualizer team used on the title page of internal specs. It actually reveals how most of us think about our product: not as a profiler (though you can get decent sample profile from it by clicking the green “Execution” category in the legend), and not even as a performance...

Visual Studio 2010 is finally here and those of us who worked on the Concurrency Visualizer are thrilled about its debut (in VS2010 Premium and Ultimate)! We truly hope that our hard work pays off for you, enabling you to solve your toughest parallel performance problems. I’d like to thank everyone who...

Those of you following the various parallel computing blogs from our team or who have played with Visual Studio 2010 have probably noticed a heavy focus on single box parallelism (and mostly on client machines). For a future version of Visual Studio we are exploring stronger investment for the...

Did you notice that reports frequently show pretty long stacks? E.g. in the picture below it is 12 frames deep, but sometimes it can be a hundred… and very frequently you need the deepest one to see what is going on. Are you supposed to click on “+” hundred times? Happily, there is a shortcut – asterisk...

The Message Passing Interface (MPI) is a popular API for developing message-passing based parallel applications on clusters. Microsoft has a Windows HPC Server product that includes an implementation of MPI, among other things (visit http://www.microsoft.com/hpc) . In this post, I’d like to demonstrate...

This post explores the performance issues that arise when using PLINQ to parallelize queries, and illustrates how the Concurrency Visualizer in Visual Studio 2010 can be a valuable tool in identifying performance bottlenecks and making efficient and profitable parallelization choices. The subject of...

One of the samples that is included with the Parallel Programming Samples for .NET 4 is a simple Ray Tracer.  This ray tracer provides a nice visual way of seeing the benefits of .NET 4 parallelism features, as well as giving insights into the way work stealing happens under the hood.  The...

We instrumented The Concurrency Visualizer within Visual Studio 2010’s profiler via Event Tracing for Windows (ETW), which depends on a number of buffers to cache data before writing it to disk. If The Concurrency Visualizer complains of lost kernel and / or user mode events during creation of a profile...

Ever wonder which of these profiling options to select?  Find out here . James Rapp - Parallel Computing Platform