mirror of
https://github.com/boostorg/compute.git
synced 2026-01-24 05:42:14 +00:00
178676df4f
This refactors the system::default_device() method. Now, the default compute device for the system is only found once and stored in a static variable. This eliminates many redundant calls to clGetPlatformIDs() and clGetDeviceIDs(). Also, the default_cpu_device() and default_gpu_device() methods have been removed and their usages replaced with default_device().
71 lines
2.4 KiB
Plaintext
71 lines
2.4 KiB
Plaintext
[section:tutorial Tutorial]
|
|
|
|
[section Hello World]
|
|
|
|
The hello world example gives a simple application that prints the name of
|
|
the default compute device on the system.
|
|
|
|
Compute devices are represented with the
|
|
[classref boost::compute::device device] class.
|
|
|
|
[import ../example/hello_world.cpp]
|
|
[hello_world_example]
|
|
|
|
[endsect] [/ hello world]
|
|
|
|
[section Transferring Data]
|
|
|
|
Before any computation occurs, data must be transferred from the host to the
|
|
compute device. The generic [funcref boost::compute::copy copy()] function
|
|
provides a simple interface for transfering data and the generic
|
|
[classref boost::compute::vector vector<T>] class provides a container for
|
|
storing data on a compute device.
|
|
|
|
The following example shows how to transfer data from an array on the host to
|
|
a [classref boost::compute::vector vector<T>] on the device and then back to
|
|
a separate `std::vector<T>` on the host. At the end of the example both
|
|
`host_array` and `host_vector` contain the same values which were copied
|
|
through the memory on the compute device.
|
|
|
|
[import ../example/copy_data.cpp]
|
|
[copy_data_example]
|
|
|
|
[endsect] [/ transferring data]
|
|
|
|
[section Transforming Data]
|
|
|
|
The following example shows how to calculate the square-root of a vector of
|
|
`float`s on a compute device using the [funcref boost::compute::transform
|
|
transform()] function.
|
|
|
|
[import ../example/transform_sqrt.cpp]
|
|
[transform_sqrt_example]
|
|
|
|
[endsect] [/ transforming data]
|
|
|
|
[section Vector Data Types]
|
|
|
|
In addition to the built-in scalar types (e.g. `int` and `float`), OpenCL
|
|
also provides vector data types (e.g. `int2` and `vector4`). These can be
|
|
used with the Boost Compute library on both the host and device.
|
|
|
|
Boost.Compute provides typedefs for these types which take the form:
|
|
`boost::compute::scalarN_` where `scalar` is a scalar data type (e.g. `int`,
|
|
`float`, `char`) and `N` is the size of the vector. Supported vector sizes
|
|
are: 2, 4, 8, and 16.
|
|
|
|
The following example shows how to transfer a set of 3D points stored as an
|
|
array of `float`s on the host the device and then calculate the sum of the
|
|
point coordinates using the [funcref boost::compute::accumulate accumulate()]
|
|
function. The sum is transferred to the host and the centroid computed by
|
|
dividing by the total number of points.
|
|
|
|
Note that even though the points are in 3D, they are stored as `float4` due to
|
|
OpenCL's alignment requirements.
|
|
|
|
[import ../example/point_centroid.cpp]
|
|
[point_centroid_example]
|
|
|
|
[endsect] [/ vector data types]
|
|
[endsect] [/ tutorial ]
|