Sycl Academy LogoSycl Academy Logo
listShow Lessons List
  1. What is SYCL
  2. Enqueueing a Kernel
  3. Managing Data
  4. Handling Errors
  5. Device Discovery
  6. Data Parallelism
  7. Introduction to USM
  8. Using USM
  9. Asynchronous Execution
  10. Data and Dependencies
  11. In Order Queue
  12. Advanced Data Flow
  13. Multiple Devices
  14. ND Range Kernels
  15. Image Convolution
  16. Coalesced Global Memory
  17. Vectors
  18. Local Memory Tiling
  19. Further Optimisations
  20. Matrix Transpose
  21. More SYCL Features
  22. Functors
Lesson
Source
Solution
GitHub Iconedit

Multiple Devices

In this exercise you will learn how to create two separate queues and then split a task across two devices.

1.) Get two devices

Get two devices. If only host device is available then both devices may be identical, that is the host device.

2.) Create queues

Then create a queue for each of the two devices.

3. ) Split a task across the two queues

Using either the application from exercise 6 using the buffer/accessor model or the application from exercise 8 with the USM model create an application where the work is distributed across two devices using the two queues you created.

Remember to create separate buffers for the separate parts of the data being computed.

Remember to wait on the work enqueued to bothqueues to complete and remember to handle errors.

Also note that you will need to have different types defined to name each of the two kernel kernel functions.

4. ) Tweak the load balancing

Try playing around with the distribution of the work between the two devices by changing the range and offset of the two buffers.

Build and execution hints

For DevCloud via JupiterLab follow these instructions.

For DPC++: instructions.

For AdaptiveCpp: instructions.