Features
Native PyTorch eager execution on TPUs
Fused Eager mode (50-100%+ speed gains)
Distributed training (DDP, FSDP)
Mixed precision training with FP8 on Ironwood TPUs
Integration with PyTorch Lightning, Hugging Face Transformers
Zero static graph compilation required
Scale to 100K+ chip clusters
Open-source backend (torch-xla) on GitHub
XLA compiler integration for optimized performance
Training and inference with vLLM on TPU
Model serving with vLLM unified backend (JAX & PyTorch)
Compatibility with existing PyTorch codebases
Supports Gemma 4 inference on vLLM TPU
Integration with MaxText for LLM training
Integration with Metrax metrics library (JAX)
Drag-and-drop low-code app builder
Pro-code with ABAP Cloud and CAP (Java/JavaScript)
Joule AI assistant for natural language app creation
AI workflow automation with reasoning and deterministic logic
Robotic process automation with UiPath add-on
AI document processing with template-based extraction
Digital workspace creation for business sites and portals
500+ prebuilt business process and app templates
Clean-core compliant extensions for S/4HANA Cloud
Integration with Visual Studio Code and MCP servers
Governance, roles, and upgrade-safe extension management
Joule Studio for building AI agents grounded in business context
Custom code migration acceleration from ABAP to S/4HANA Cloud
Generative AI-based code development for Java/JavaScript on CAP
ABAP AI capabilities for cloud-ready business applications
