Back to News
PyTorch (Meta / Linux Foundation)•May 13, 2026
PyTorch 2.12
Paid
Deep Learning
PyTorch
CUDA
GPU Computing
Framework
Open Source
### TL;DR
PyTorch 2.12 brings major performance improvements including up to 100x faster batched linear algebra (linalg.eigh) on CUDA via updated cuSOLVER API selection, the new torch.accelerator.Graph API for optimized hardware execution, experimental CUDA 13.2 support, and broad updates across compilation, distributed systems, export, and graph capture.
Key Insights & Metrics
Pricing
Free (Open Source)
Cost structure
Version
2.12
Current release version
Hardware
CUDA-compatible GPU for GPU acceleration; CPU-only mode also supported
Compute requirements
Category
Paid
Licensing model
Region
United States
Primary region
Key Features
- Up to 100x faster batched linalg.eigh on CUDA via cuSOLVER backend update
- New torch.accelerator.Graph API for optimized hardware execution
- Experimental CUDA 13.2 support and updates across compilation and distributed systems
Discussion
0
Upvotes
0
Downvotes
0 reviews
Sign in to leave a review
Reviews
No reviews yet. Be the first to review!