Skip to content

Latest commit

 

History

History
89 lines (64 loc) · 3.22 KB

README.md

File metadata and controls

89 lines (64 loc) · 3.22 KB
Raw

SYCL Academy

Exercise 2: Hello World

In this first exercise you will learn how to enqueue your first kernel function to run on a device and print Hello World! to the console.

1.) Create a queue

The first thing you must do is create a queue to submit work to. The simplest way to do this is to default construct it, this will choose a device for you.

2.) Define a command group

Once you have a queue you can now submit work to be executed on the device that queue is targeting and this is done via a command group.

Define a lambda expression which takes a reference to a handler to represent your command group function and pass it to the submit member function of the queue.

Note that submitting a command group without any commands will result in an error.

3.) Define a SYCL kernel function

Within the command group function define a SYCL kernel function via the single_task command within the command group, which takes only a function object which itself doesn't take any parameters.

Remember to declare a class for your kernel name in the global namespace.

Also remember to call wait on the event returned from submit to await the completion of the kernel function.

4.) Stream "Hello World!" to stdout

Create a stream object within the command group scope. The parameters to the constructor of the stream class are the total buffer size, the work-item buffer size and the handler.

Then use the stream you constructed within the SYCL kernel function to print "Hello world!" using the << operator.

Build And Execution Hints

For For DPC++ (using the Intel DevCloud):

clang++ -fsycl -o sycl-ex-2 -I../External/Catch2/single_include ../Code_Exercises/Exercise_02_Hello_World/source.cpp

In Intel DevCloud, to run computational applications, you will submit jobs to a queue for execution on compute nodes, especially some features like longer walltime and multi-node computation is only abvailable through the job queue. Please refer to the guide.

So wrap the binary into a script job_submission and run:

qsub job_submission

For ComputeCpp:

cmake -DSYCL_ACADEMY_USE_COMPUTECPP=ON -DSYCL_ACADEMY_INSTALL_ROOT=/insert/path/to/computecpp ..
make exercise_02_hello_world_source
./Code_Exercises/Exercise_02_Hello_World/exercise_02_hello_world_source

For hipSYCL:

# <target specification> is a list of backends and devices to target, for example
# "omp;hip:gfx900,gfx906" compiles for CPUs with the OpenMP backend and for AMD Vega 10 (gfx900) and Vega 20 (gfx906) GPUs using the HIP backend.
# The simplest target specification is "omp" which compiles for CPUs using the OpenMP backend.
cmake -DSYCL_ACADEMY_USE_HIPSYCL=ON -DSYCL_ACADEMY_INSTALL_ROOT=/insert/path/to/hipsycl -DHIPSYCL_TARGETS="<target specification>" ..
make exercise_02_hello_world_source
./Code_Exercises/Exercise_02_Hello_World/exercise_02_hello_world_source

alternatively, without cmake:

cd Code_Exercises/Exercise_02_Hello_World
/path/to/hipsycl/bin/syclcc -o sycl-ex-2 -I../../External/Catch2/single_include --hipsycl-targets="<target specification>" source.cpp
./sycl-ex-2