# Search:

Type: Posts; User: Rui

Page 1 of 2 1

1. ## Thread: OpenCL ideas for Masters Dissertation

by Rui
Replies
1
Views
1,884

### OpenCL ideas for Masters Dissertation

Hello,

In a couple of months I'm about to start my Masters Dissertation and I'm looking for some ideas. Since I have some background with OpenCL (and some Cuda knowledge as well), I'm looking for...
2. ## Thread: Difference between USE_HOST, ALLOC_HOST, COPY_HOST pointer

by Rui
Replies
2
Views
5,529

### Difference between USE_HOST, ALLOC_HOST, COPY_HOST pointer

What are the major differences between those three? Can someone provide examples if necessary?
3. ## Thread: Limitation of local sizes on N-Dimensional NDRanged problems

by Rui
Replies
1
Views
1,867

### Limitation of local sizes on N-Dimensional NDRanged problems

If my graphics card has a limitation of 512 work-items per work-group and I'm modelling a 2-dimensional problem, that means the multiplication of the local size I chose for each dimension can be no...
4. ## Thread: Double precision floats and built in math functions?

by Rui
Replies
4
Views
3,882

### Re: Double precision floats and built in math functions?

Not a bug. You are using pow function wrongly. From the spec:

gentype pow (gentype x, gentype y)
gentype pown (gentype x, intn y)

See what's wrong? You have pow(4, 2) where you should have...

by Rui
Replies
0
Views
2,105

### Parallel primitives

I'm looking for parallel primitives that I can apply on the problem I'm solving, because I'm lacking some theoretical concepts. I've already been told about reduce, can anyone tell about existing...

by Rui
Replies
1
Views
2,030

What is the best way (as in the most efficient) to calculate the sum of a big array (>80000 cells)? In the problem I'm trying to solve I have at least 100 work-items, each one calculating the sum of...

by Rui
Replies
1
Views
1,980

### Performance tips

What are the general tips/tricks to improve kernel performance? Something like try to use constant instead of global when possible, etc, etc? Can anyone name a few, based on his experience?
8. ## Thread: Performance 1 .cl file vs several .cl files

by Rui
Replies
1
Views
1,951

### Performance 1 .cl file vs several .cl files

In what concerns to performance, is it different to use only one .cl file with several kernels instead of for example one kernel on its own file?

by Rui
Replies
1
Views
3,706

### Re: using opencl link error..

You've done it this way http://www.khronos.org/message_boards/viewtopic.php?f=37&t=2651&p=6969&hilit=visual+studio#p6969 and you're using the right card drivers?
10. ## Thread: Compile OpenCL code to binary

by Rui
Replies
2
Views
3,402

### Re: Compile OpenCL code to binary

Is there any example of this available?
11. ## Thread: Printing a cl_mem object

by Rui
Replies
1
Views
3,099

### Re: Printing a cl_mem object

If you are talking about a buffer, you have to enqueue a read to that buffer and place the content in an array on host.
12. ## Thread: Compile OpenCL code to binary

by Rui
Replies
2
Views
3,402

### Compile OpenCL code to binary

Since there is a function clCreateProgramWithBinary, instead of compiling code on runtime, is it possible to compile OpenCL code to binary format, officially speaking?
13. ## Thread: User defined functions in OpenCL code

by Rui
Replies
1
Views
2,391

### User defined functions in OpenCL code

Is it possible to create our own functions in OpenCL code, so that the kernels could call them? Where can I see some simple example of doing this?
14. ## Thread: How can I simulate double precision?

by Rui
Replies
2
Views
2,850

### How can I simulate double precision?

I'm solving a problem that uses double precision on C++, but double precision is not supported by all hardware, specially the hardware I'm currently working on. In that case, how can I simulate the...
15. ## Thread: event return by EnqueueNDRange broken with nVidia

by Rui
Replies
5
Views
4,247

### Re: event return by EnqueueNDRange broken with nVidia

I'm experiencing some trouble with that too.
16. ## Thread: What is error -5 : out of resource

by Rui
Replies
3
Views
2,788

### Re: What is error -5 : out of resource

For example, imagine you are working with a graphics card with 512Mb of memory, and you're trying to pass to your kernel a cl_int array with size of 50000*10000. Making some calculations.. you're...
17. ## Thread: Mysterious kernel behavior. Can anyone explain?

by Rui
Replies
7
Views
3,891

### Re: Mysterious kernel behavior. Can anyone explain?

I experienced a similar problem some time ago when executing nested for loops on the kernel, but couldn't find at the time any exact explanation for it.
18. ## Thread: Error -30 invalid value...

by Rui
Replies
1
Views
2,109

### Error -30 invalid value...

What could be the cause of this error? All other error description give some sort of explanation about its cause, but this one is tricky... What situations may cause this one?

by Rui
Replies
1
Views
2,007

### Effective gain by using padding on workgroup size

Let's say I have several work groups with size 349. According to what I read, it's better to use a work group size of 2^n instead of for example those 349. Can I expect a considerable gain of...
20. ## Thread: Error enqueueing kernel

by Rui
Replies
1
Views
1,976

### Re: Error enqueueing kernel

Never mind, keeping OpenCL stuff on a separate class is OK. Buffer creation was buggy.
21. ## Thread: Error enqueueing kernel

by Rui
Replies
1
Views
1,976

### Error enqueueing kernel

Hello,

When I try to enqueue a kernel, I'm getting an error saying "access violation reading location 0xfdfdfdfd". I've put most of the OpenCL related stuff on a class (context creation, building,...

by Rui
Replies
3
Views
3,339

According to the specification there is a function
int atom_add (__local int *p, int val)
23. ## Thread: Problem with first code on visual studio 2008

by Rui
Replies
4
Views
3,656

### Re: Problem with first code on visual studio 2008

No problem, glad to help =)
24. ## Thread: how to debug opencl kernel program running on GPU?

by Rui
Replies
7
Views
12,412

### Re: how to debug opencl kernel program running on GPU?

Well, my technique to detect syntax errors is to comment code and run it to check if it builds ok. To see your results you can put them in a buffer and check it outside the kernel.
25. ## Thread: Problem with first code on visual studio 2008

by Rui
Replies
4
Views
3,656

### Re: Problem with first code on visual studio 2008

You have to add the lib file to the linker input files. It should be on the folders of the OpenCL SDK.