Will be AMP GPU debugging supported in Windows 7?
In current CTP, debugging of GPU code is not supported for Windows 7. Is this limitation only for CTP or even the final C++ AMP will require Windows 8 to debug on GPU? ThanksMartin Pozor
View Articlewarning lnk4006
I did link in my program to WINTRUST.dll and also to CRYPT32.dll, so I get a lot of warnings that fanctions already defined. there is a lot of functions that defined twice- is those two dll's. What...
View Articledebugging of MPI code in visual c++ 2012
I have installed MS-MPI and linked with Visual Studio 2012 (C++) and made a simple MPI base parallel program in C++ to run and I succeeded. But I can not debug this program. How Can I debug this...
View ArticleDeploying in Win7 computer
I have developed an C++ AMP application and C++ AMP dll on Win8 computer with Visual Studio 2012. I have tested both and they work fine. After I deployed them in a Win7 computer, it crashes inside AMP....
View ArticleI don't find any data copy performance improvements since beta
Hi, I have a hobby project where I try to implement neural network based computations using GPGPU programming. I have an OpenCL layer which performs really well, but I want to implement a C++ AMP based...
View ArticleFunction overloading with restrict in C++ AMP and namespaces
So I read the article here http://blogs.msdn.com/b/nativeconcurrency/archive/2012/03/29/function-overloading-with-restrict-in-c-amp.aspx that was very useful. However I had an issue with overloading...
View ArticleHelloWorldCSharpWinRT won't compile
Hello, I am trying to play with winRT cpp amp demo found here: http://blogs.msdn.com/b/pfxteam/archive/2011/11/11/how-to-use-c-amp-from-c-using-winrt.aspx When I try to compile I get this error:...
View ArticleAny optimization ways in memory transfer in AMP?
Recently we did some research on Image processing using GPGPU, and we tired both OpenCL and AMP. It was found that AMP performs worse than OpenCL, consuming more time with the same workload. And we...
View ArticleRecommended way to handle I/O in PPL tasks
I'm trying to figure out what the best practice is to handle potentially blocking I/O calls in code that's using PPL tasks. Based on my current understanding I see a few options but it is not clear to...
View ArticleError when synchronizing data from GPU to CPU
Hi, Recently I have a weird problem with synchronizing array_views from GPU to CPU memory. The short example below demonstrates this: #include <vector> #include <amp.h> #include...
View ArticlePPL assigning a Scheduler to a task_group member variable
Hi, I'm actually having difficulty with this. class Foo has a private member "task_group _tasks"... how do I properly assign a scheduler to this task_group that is NOT the default scheduler? My...
View Articlechange a exe application code into dll and used in C++ app
Hello, Can any one help me on below: I have a c exe application source code (project file pack) and I need to port into my C++ application, how can I change the project setting to may it as DLL, so I...
View ArticleAMP fatal error C1001
The following code generates a fatal error. parallel_for_each(extent<1>(sample_size).tile<ggsc_tile_size>(), [=] (tiled_index<ggsc_tile_size>tidx)restrict(amp) { ... if(tidx.local[0]...
View ArticleLog barrier in posix
I've been trying to implement logarithm barrier with condition variables using this algorithm (found on dna.cs.byu.edu/CS484/info/log-barrier-using-pthread-condition-variables) typedef struct...
View Articleconcurrent_unordered_map samples pack - very slow performance under debug - VS10
Hi, I use concurrent_unordered_map and saw that under debug mode while adding values to the container things move very slow compared to the same scenario under release mode. To give you a measure of...
View ArticleDynamic code generation and AMP
Just some feedback. I am a PhD. student and my thesis topic is primarily about developing a dynamic compilation framework that generates GPU code on the fly. While I love the design of AMP,...
View Articletile static memory use
tile_static memory can be used only within a tile parallel_for_each. In my case, I want to use tile_static memory, but I do not need to use tiled parallel_for_each. I tried a workaround by having a...
View Articleerror MSB8008 , C++AMP VS12 on Windows 7
hello Error 1 error MSB8008: Specified platform toolset (v110) is not installed or invalid. Please make sure that a supported PlatformToolset value is selected. C:\Program Files...
View ArticleDo simultaneous AMP calls need locks?
I'm P/Invoking into AMP from a C# project and I have a question regarding multiple simultaneous calls to AMP kernels. If I have more than 1 thread making calls into AMP at a time, do I need to put in...
View ArticleAMP Performance in 64 bits
Hi,After getting a new machine, I tested a bit more my AMP code, and I get some strange result between a 32 and 64 bits binary.I tested on a freshly installed Windows 7 Pro 64 bits SP1. I use only...
View Article