Developer Resources for Consumer Internet

NVIDIA® CUDA-X AI™ is a complete deep learning software stack for researchers and software developers to build high-performance, GPU-accelerated applications for conversational AI, recommendation systems, and computer vision. CUDA-X AI libraries deliver world-leading performance for both training and inference across industry benchmarks such as MLPerf.

NVIDIA TensorRT™ is an SDK for high-performance deep learning inference. It includes a deep learning inference optimizer and runtime that delivers low latency and high throughput for deep learning inference applications.

NVIDIA RAPIDS™ is an open-source suite of data processing and machine learning libraries, developed by NVIDIA, that enables GPU acceleration for data science workflows. RAPIDS relies on NVIDIA’s CUDA® language, allowing users to leverage GPU processing and high-bandwidth GPU memory through user-friendly Python interfaces.

GPU-accelerated Apache Spark 3.0 speeds up data science pipelines—without code changes—and data processing and model training while substantially lowering infrastructure costs.

The Deep Learning Profiler is a tool for profiling deep learning models to understand and improve performance of data science models visually via TensorBoard or by analyzing text reports.

NVIDIA Nsight™ Systems is a system-wide performance analysis tool designed to visualize an application’s algorithms, to help you identify the largest opportunities for optimiz and tuning to scale efficiently across any quantity or size of CPUs and GPUs, from large servers to the smallest system on a chip (SoC).

NGC provides a range of options that meet the needs of data scientists, developers, and researchers with various levels of AI expertise. Quickly deploy AI frameworks with containers, get a head start with pre-trained models or model training scripts, and use domain-specific workflows and Helm charts for the fastest AI implementations, giving you faster time to solution.

Kubernetes on NVIDIA GPUs enables enterprises to scale up training and inference deployment to multi-cloud GPU clusters seamlessly. It lets you automate the deployment, maintenance, scheduling, and operation of multiple GPU-accelerated application containers across clusters of nodes.

The NVIDIA Data Center GPU Manager (DCGM) is a set of tools for managing and monitoring NVIDIA GPUs in cluster environments. It's a low-overhead tool suite that performs a variety of functions on each host system, including active health monitoring, diagnostics, system validation, policies, power and clock management, group configuration, and accounting.

Data collected on a vast scale has fundamentally changed the way organizations do business, driving demand for teams to provide meaningful data science, machine learning, and deep learning-based business insights quickly. Learn how data science leaders can use RAPIDS to boost their teams’ productivity while optimizing their costs and minimizing deployment time.

Real-time conversational AI is a complex and challenging task. Explore NVIDIA Riva and how to access its high-performance conversational AI services easily and quickly with just a few commands.

This blog describes a highly optimized, GPU-accelerated inference implementation of a wide-and-deep model based on TensorFlow’s DNNLinearCombinedClassifier API. The proposed solution allows for easy conversion from a trained TensorFlow wide-and-deep model to a mixed-precision inference deployment.

NVIDIA has worked with the Apache Spark community to implement GPU acceleration with the release of Spark 3.0 and the open-source RAPIDS Accelerator for Spark. In this blog, learn how the RAPIDS Accelerator for Apache Spark uses GPUs to speed up end-to-end data preparation and model training on the same Spark cluster, Spark SQL, and DataFrame operations without requiring any code changes.

BERT (Bidirectional Encoder Representations from Transformers) provides a game-changing twist to the field of natural language processing (NLP). It runs on supercomputers powered by NVIDIA GPUs to train its huge neural networks and achieve unprecedented NLP accuracy, impinging in the space of known human language understanding. AI like this has been anticipated for many decades. With BERT, it’s finally arrived.

Click-through rate (CTR) estimation is one of the most critical components of modern recommender systems. In this blog, get an introduction to HugeCTR, a GPU-accelerated training framework for CTR estimation and a pillar of NVIDIA Merlin. HugeCTR, on a single NVIDIA V100 Tensor Core GPU, achieves a speedup of up to 114X over TensorFlow on a 40-core CPU node and up to 8.3X that of TensorFlow on the same V100 GPU.