TornadoVM P2Code Runner (NVIDIA)

This project provides a Docker-based environment to compile and execute TornadoVM-accelerated Java programs using the OpenCL backend on NVIDIA GPUs.

---

1. Run an Example

You can run the VectorAddInt.java class which uses the TornadoVM API to offload the execution of a method for execution on NVIDIA GPUs. By invoking the run-software-acceleration.sh script, you will pull the latest docker image and will run the example as follows.

./run-software-acceleration.sh examples/VectorAddInt.java
🔧 Compiling /data/examples/VectorAddInt.java...
✅ Compilation successful
🚀 Running with TornadoVM: VectorAddInt
WARNING: Using incubator modules: jdk.incubator.vector
Task info: s0.t0
    Backend           : OPENCL
    Device            : NVIDIA RTX A2000 8GB Laptop GPU CL_DEVICE_TYPE_GPU (available)
    Dims              : 1
    Global work offset: [0]
    Global work size  : [256]
    Local  work size  : [256, 1, 1]
    Number of workgroups  : [1]

Result is correct

2. Extend the Dockerfile and build the Docker Image

Make sure you have Docker installed and NVIDIA drivers properly set up for GPU passthrough.

• To build the Docker image, run:

./buildDocker.sh