Use Cases

Real-world applications

Deploy ML models on edge devices

Compile and optimize machine learning models to run efficiently on resource-constrained edge hardware using minimal runtimes.

Optimize ML workloads for data center GPUs

Customize compilation pipelines to maximize performance of ML models on high-performance GPUs in data centers.

Cross-platform ML model deployment

Use universal deployment modules to run the same ML model across different hardware architectures without rewriting code.