TornadoVM

TornadoVM copied to clipboard

[codegen][opencl] issue with codegen on complex control-flow in opencl

1

### Describe the bug There is a bug with opencl code gen with complex control flow and the kernel context api: ```java public static void calculateAttentionScores(KernelContext context, IntArray positionNlayer, int...

mikepapadim

Support for JDK 24 and migrate to Graal compiler 24.2.0

2

This PR is **WIP** for TornadoVM running with JDK24 and GraalVM 24.0.0. How to test: GraalVM version used for initial migration: ```bash wget https://download.oracle.com/graalvm/24/latest/graalvm-jdk-24_linux-x64_bin.tar.gz ``` ```bash make graal-jdk-24 BACKEND=opencl make...

mikepapadim

### Describe the bug In `macOS`, I noticed that the `TestAtomics` unit-tests have one test failing. The error is related to the `cl2Metal` conversion which implies that the Metal clang...

stratika

Unsupported type error for int and uint types

1

### Description Running TornadoVM powered applications on a PC with Turkish locale yields to following error: ```text Unable to compile task graph.double - computeDouble The internal error is: [Error during...

selcuksert

auto quantization of big arrays on computation

1

hi , is it possible to auto quantization of big arrays on long computations and make them back to their original values. Maybe we can use auto encoders or gaussians...

eix128

[TensorAPI] Add initial support for TensorQ8 and TensorQ4 to support quantization and dequantization of Floats

2

#### Description ***NO CODE GEN SUPPORT*** This PR adds support for Q8 and Q4 tensor quantization in TornadoVM. The implementation includes: * A new TensorQ8 class that implements 8-bit quantization...

mikepapadim