# CUDA Framework Guides

Master GPU acceleration with 20+ framework guides. Each guide includes installation, code examples, performance tips, and CUDA optimization strategies.

## Deep Learning Frameworks

[**PyTorch** Python  
Complete PyTorch CUDA guide. Learn GPU memory management, mixed precision training, custom CUDA kernels, and performance optimization for deep learning.](/content/guides/frameworks/pytorch/index.html)  
[**TensorFlow** Python  
Complete TensorFlow CUDA guide. Learn XLA compilation, mixed precision, memory growth, distributed training, and TensorRT optimization.](/content/guides/frameworks/tensorflow/index.html)  
[**JAX** Python  
Complete JAX CUDA guide. Learn JIT compilation, automatic vectorization, parallelization with pmap, and XLA optimization for GPU computing.](/content/guides/frameworks/jax/index.html)  
[**Horovod** Python  
Complete Horovod guide. Learn distributed training with PyTorch, TensorFlow across multiple GPUs.](/content/guides/frameworks/horovod/index.html)  
[**DeepSpeed** Python  
Complete DeepSpeed guide. Learn ZeRO optimization, large model training, and inference.](/content/guides/frameworks/deepspeed/index.html)  
[**Megatron** Python  
Complete Megatron-LM guide. Learn tensor and pipeline parallelism for training GPT/BERT at scale.](/content/guides/frameworks/megatron/index.html)  
[**Apex** Python  
Complete NVIDIA Apex guide. Learn AMP, distributed training utilities, and fused optimizers.](/content/guides/frameworks/apex/index.html)

## Scientific Computing Frameworks

[**CuPy** Python  
Complete CuPy guide for GPU-accelerated NumPy. Learn array operations, custom kernels, interoperability with PyTorch/TensorFlow, and CUDA optimization.](/content/guides/frameworks/cupy/index.html)  
[**RAPIDS** Python  
Complete RAPIDS guide for GPU data science. Learn cuDF for DataFrames, cuML for machine learning, cuGraph for graph analytics, and CUDA optimization.](/content/guides/frameworks/rapids/index.html)  
[**PyCUDA** Python  
Complete PyCUDA guide. Learn to write CUDA C kernels and call them from Python.](/content/guides/frameworks/pycuda/index.html)

## CUDA Compilers & Code Generation

[**Triton** Python  
Complete Triton guide for GPU kernel development. Learn to write custom CUDA kernels in Python with automatic optimization and PyTorch integration.](/content/guides/frameworks/triton/index.html)  
[**Numba** Python  
Complete Numba CUDA guide. Learn JIT compilation, CUDA kernels in Python, GPU arrays, vectorization, and performance optimization for scientific computing.](/content/guides/frameworks/numba/index.html)  
[**TVM** Python/C++  
Complete TVM guide for ML compilation. Learn AutoTVM tuning, Relay IR, CUDA code generation, operator fusion, and deployment optimization.](/content/guides/frameworks/tvm/index.html)  
[**Numba** Python  
Complete Numba CUDA guide. Learn JIT compilation, GPU kernels in Python, and performance optimization.](/content/guides/frameworks/numba/index.html)

## GPU Acceleration Libraries

[**TensorRT** Python/C++  
Complete TensorRT guide for optimized inference. Learn layer fusion, precision calibration, dynamic shapes, plugins, and deployment on NVIDIA GPUs.](/content/guides/frameworks/tensorrt/index.html)  
[**ONNX Runtime** Python/C++  
Complete ONNX Runtime guide for GPU inference. Learn CUDA execution providers, graph optimizations, quantization, and deployment across platforms.](/content/guides/frameworks/onnxruntime/index.html)  
[**RAPIDS** Python  
Complete RAPIDS guide. Learn cuDF, cuML, cuGraph for GPU-accelerated data science.](/content/guides/frameworks/rapids/index.html)  
[**ORT** Python/C++  
Complete ONNX Runtime CUDA guide. Learn model optimization, TensorRT integration, and deployment.](/content/guides/frameworks/onnxruntime/index.html)  
[**TRT** Python/C++  
Complete TensorRT guide. Learn model optimization, quantization, and deployment for NVIDIA GPUs.](/content/guides/frameworks/tensorrt/index.html)  
[**cuDF** Python  
Complete cuDF guide. Learn GPU-accelerated DataFrame operations with pandas API.](/content/guides/frameworks/cudf/index.html)

## Optimize Any Framework Automatically

RightNow AI integrates with PyTorch, TensorFlow, JAX, and more to provide real-time CUDA optimization.

[Download RightNow AI→](/content/downloads/index.html)