about reference points
opened this issue · comments
This code is helpful for beginners learning cuda, why we can't increase the limit...
I used cudaMallocPitch for optimal performances (compared to cudaMalloc).
The problem with this allocation is that you cannot allocate too many
points.
I don't know if this still applies with newer CUDA versions.
Check the documentation of cudaMallocPitch for more information.
Okay thanks i will check.
…On 3 Aug 2017 9:13 p.m., "Vincent Garcia" ***@***.***> wrote:
I used cudaMallocPitch for optimal performances (compared to cudaMalloc).
The problem with this allocation is that you cannot allocate too many
points.
I don't know if this still applies with newer CUDA versions.
Check the documentation of cudaMallocPitch for more information.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#3 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/Ab6fFKDbz39N2KcNABYk9SlS2iEmGAmRks5sUfG4gaJpZM4OsAvr>
.
@vincentfpgarcia Sire can you please describe me this term "#define MAX_PART_OF_FREE_MEMORY_USED 0.9" why used. what value i can update according to newer specification of texture width, height[65536,65536] not older one.
#define MAX_PITCH_VALUE_IN_BYTES = 262144 =>can replaced by Max 2D Linear Texture Pitch: 1048544 bytes/ Max Memory Pitch: 2147483647 bytes ????which one used how you decide to use 262144 this value
#define MAX_TEXTURE_WIDTH_IN_BYTES 65536 => same width
#define MAX_TEXTURE_HEIGHT_IN_BYTES 32768 => Max 2D Texture Size: 65536 x 65536
#define MAX_PART_OF_FREE_MEMORY_USED 0.9 ? how you decide it
#define BLOCK_DIM 16
MAX_PART_OF_FREE_MEMORY_USED defines how much of the free memory available
the program is going to use for the computation. I wanted to make sure that
not all graphic memory is used by the program in order to avoid problems.
As I understand, you are using this code for your master project right ?
Please tell me what is your school, the master you're in and the name of
your professor(s) involved in this project.
Thank you.
…On Mon, Aug 7, 2017 at 11:21 AM, ayshaashraf9002 ***@***.***> wrote:
@vincentfpgarcia <https://github.com/vincentfpgarcia> can you please
describe me this term why used. what value i can update according to
texture width, height. as according to cuda compute compatibility 5.0
#define MAX_PART_OF_FREE_MEMORY_USED 0.9
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#3 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ADbwrUVzHNta9_ZgeBab444B8n5cKYhcks5sVtcVgaJpZM4OsAvr>
.
Sir thanks 4 replying. i am learning the concepts of parallel processing
using cuda and mex file. Your work is bit helpful thatswhy i am asking all
these questions.
Sir i want to ask about openCl which one is more good in your point
of view. In CUDA i practice thrust library also. Its very good to use.
GitHub is not a forum. I've already answered many of your question.
You haven't answered mine. What is you school? Give me the name of your
advisors and a way to contact them.
On Tue, Aug 8, 2017 at 6:18 PM, ayshaashraf9002 <notifications@github.com>
wrote:
… Sir i want to ask about openCl or cuda which one is more good in your point
of view. In cuda i practice thrust library also. Its very good to use.
On 8 Aug 2017 9:00 p.m., "Ay-sha Ashraf" ***@***.***> wrote:
Sir thanks 4 replying. i am learning the concepts of parallel processing
using cuda and mex file. Your work is bit helpful thatswhy i am asking all
these questions.
On 8 Aug 2017 8:11 p.m., "Vincent Garcia" ***@***.***>
wrote:
> MAX_PART_OF_FREE_MEMORY_USED defines how much of the free memory
available
> the program is going to use for the computation. I wanted to make sure
that
> not all graphic memory is used by the program in order to avoid problems.
>
> As I understand, you are using this code for your master project right ?
> Please tell me what is your school, the master you're in and the name of
> your professor(s) involved in this project.
> Thank you.
>
>
> On Mon, Aug 7, 2017 at 11:21 AM, ayshaashraf9002 <
***@***.***
> >
> wrote:
>
> > @vincentfpgarcia <https://github.com/vincentfpgarcia> can you please
> > describe me this term why used. what value i can update according to
> > texture width, height. as according to cuda compute compatibility 5.0
> > #define MAX_PART_OF_FREE_MEMORY_USED 0.9
> >
> > —
> > You are receiving this because you were mentioned.
> > Reply to this email directly, view it on GitHub
> > <#3#
> issuecomment-320614262>,
> > or mute the thread
> > <https://github.com/notifications/unsubscribe-auth/
> ADbwrUVzHNta9_ZgeBab444B8n5cKYhcks5sVtcVgaJpZM4OsAvr>
> > .
> >
>
> —
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub
> <#3 (comment)-
320986769>,
> or mute the thread
> <https://github.com/notifications/unsubscribe-auth/Ab6fFEbPe-
rvPQXvjMAcGhzvFtyH93gCks5sWHqxgaJpZM4OsAvr>
> .
>
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#3 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ADbwrSTtC54ib2VEy77SI8oRH5IUx-8iks5sWIo-gaJpZM4OsAvr>
.
Sir my work is different from this work. I am not using this for project.
Sorry for asking questions about open Cl and thurst here.
I get the following errors:
MEMORY ALLOCATION ERROR: unspecified launch time failure
Whished allocated memory type Error