Building AI? NGC Sessions at GTC Show How to Accelerate AI Development

by Chintan Patel

From e-commerce to healthcare to agriculture, enterprises across virtually every industry are adopting AI because they know it can help them better serve customers, improve operational efficiency and gain a competitive edge.

However, building an AI solution is not always straightforward, and data scientists and developers can greatly benefit from tools that simplify their efforts.

NGC, NVIDIA’s hub of GPU-optimized AI software, is built to do just that: simplify and accelerate AI development.

And at GTC, running April 12-16, experts will present a variety of technical talks and Q&A panels to help people new to AI or those just looking for tools to speed up their AI development using the various components of NGC, including:

  • AI containers optimized to speed up AI/ML training and inference,
  • Pre-trained models that provide an advanced starting point to build custom models,
  • Industry-specific AI software development kits that transform applications into AI-powered ones,
  • Helm charts to provide consistent and faster deployments, and
  • Collections that bring together all the software needed for various use cases.

Attend the GTC talks below for walkthroughs on how to build solutions with NVIDIA AI software from NGC. And join the live Q&A sessions to have experts answer your questions related to NGC and container security.

Accelerating AI Workflows with NVIDIA NGC

This session will walk through building a conversational AI solution using the artifacts from the NGC catalog, including a Jupyter notebook, so the process can be repeated offline. It will also cover the benefits of using NGC software throughout AI development journeys.

Session time: 8 a.m. PT on Tuesday, April 13. Register here.

Building a Text-to-Speech Service that Sounds Like You

This session will build a TTS model for expressive speech using pre-trained models developed by NVIDIA Research with NVIDIA AI software. The model will be fine-tuned with speech samples and customized for the variability in speech performing style transfer from other speakers. The provided tools let developers create a model for their voice and style and make the TTS service sound just like them!

Register here.

Analyzing Traffic Video Streams at Scale

This session will demonstrate how to use the TAO Toolkit and pre-trained models to build computer vision models and run inference on over 1,000 live video feeds on a single AWS instance powered by NVIDIA A100 GPUs.

Register here.

Industry Experts Discuss Container Security and Best Practices

Security is a top concern for enterprises, and it’s even more critical with container deployments going mainstream. A panel of security experts will cover common pitfalls developers should watch for during development, security trends in the data center, cloud and hybrid, and how DevSecOps and IT teams can encourage developers to integrate security during development to ensure secure and compliant applications.

Register here.

Meet NVIDIA Experts for Deep Learning, Machine Learning and HPC

This “Connect with the Experts” session lets developers meet one-on-one with NVIDIA engineers to get their NGC-related questions answered. It’ll focus on using GPU-accelerated software from NGC as well as pre-trained models and industry SDKs with Docker, Singularity and Kubernetes. Data scientists, developers, devops and system admins supporting deep learning, machine learning, data analytics and HPC workloads will benefit the most.

Session time: 11 a.m. PT on Monday, April 19. Register here.

Deploy Compute and Software Resources to Run State-of-the-Art GPU-Supported AI/ML Applications in Azure Machine Learning with Just Two Commands

This session will demonstrate building a taxi fare prediction application using RAPIDS and show how to automatically set up a DASK cluster with multiple Azure virtual machines to support large datasets, mount data into the Dask scheduler and workers, deploy GPU-optimized AI software from NGC to train models, and then make taxi fare predictions.

Register here.

Overcoming HPC Application Communication Bottlenecks with Intelligent and Automatic Resource Selection

Performance in many HPC applications can suffer due to poor GPU-to-GPU communication stemming from inefficient resource selection when running on multi-GPU systems. This session introduces a novel tool, called NVTAGS, that suggests efficient GPU assignments and CPU/NIC affinity settings to speed up application performance.

Register here.

Building Blocks for a Successful AI-Driven Product Cycle

Learn how Lightricks, the maker of the award-winning app called FaceTune, is bringing state-of-the-art AI solutions to its millions of customers on a regular basis and how efficient ML frameworks and pipelines make the work of its research department count. Lightricks will also show how to efficiently use compute resources with a focus on mixed on-prem and multi-cloud NVIDIA GPUs and how to utilize the software from providers such as NGC.

Register here.

And don’t miss GTC’s opening keynote address from NVIDIA CEO Jensen Huang, on April 12 at 8:30 a.m. Pacific time.