ICYMI: NVIDIA TensorRT and Triton in Healthcare

In Case You Missed It (ICYMI) is a series in which we feature key lectures, white papers, blogs, and success stories that showcase NVIDIA technologies that accelerate real-world solutions.

In this update, we look at how NVIDIA TensorRT and the Triton Inference Server can help your business deploy high-performance, resilient models on a large scale. We start with a detailed, step-by-step introduction to TensorRT and Triton. Next, let’s examine exactly how Triton and Clara Deploy complement each other in your healthcare use cases. To top it off, our whitepaper covers exactly what you need to know when migrating your applications to Triton.

TensorRT and Triton in practice

On demand: Inception Café – Accelerate deep learning inference with NVIDIA TensorRT and Triton
A step-by-step guide to using NVIDIA TensorRT and Triton in conjunction with NVIDIA Clara Deploy.

Whitepaper: Inception Café – Migrating Your Medical AI App to Triton
This white paper explores the end-to-end process of migrating an existing medical AI application to Triton.

On-Demand: Introduction to TensorRT and Triton A walkthrough to optimize your first deep learning inference model
An overview of TensorRT optimization of a PyTorch model, followed by deploying the optimized model with Triton. At the end of this workshop, developers will see the significant benefits of integrating TensorRT and begin optimizing their own deep learning models.

Clara imaging

On-Demand: Clara Train 4.0 – 101 First Steps
This session provides a tour of the features and capabilities of the Clara Train SDK using a range of Jupyter notebooks covering a range of topics including Medical Model Archives (MMARs), AI-assisted annotation, and AutoML. These features help data scientists quickly annotate, train, and optimize hyperparameters for their deep learning model.

On-Demand: Clara Train 4.0 – 201 Federated Learning
This session provides an overview of federated learning, a distributed AI model development technique that can be used to build models without transferring data outside of hospitals or imaging centers. The session ends with a walkthrough of Clara Train’s federated learning capabilities by going through a series of Jupyter notebooks.

On demand: Medical imaging AI with MONAI Bootcamp
MONAI is a freely available, community-supported, open source framework based on PyTorch for deep learning in medical imaging. It provides domain-optimized basic functions for developing training workflows for medical imaging in a native PyTorch paradigm. This MONAI boot camp offers medical imaging researchers an in-depth look at the architecture of MONAI and ends with an overview of MONAI’s capabilities using a set of four Jupyter notebooks.

Clara Guardian

On-Demand: Clara Guardian 101: A Hello World Walkthrough on the Jetson Platform
NVIDIA Clara Guardian provides pre-trained models and sample applications for healthcare that can dramatically reduce the time to solution for developers building intelligent hospital applications. It targets three categories: public safety (thermal screening, mask detection and social distancing monitoring), patient care (patient monitoring, fall detection and patient involvement), and operational efficiency (operating room workflow automation, surgical analysis and contactless control). . In this session, attendees will get a walkthrough of using Clara Guardian on the Jetson NX platform, including using the pre-trained models for tasks such as automatic speech recognition and posture estimation.

Clara Parabricks

On Demand: GPU Accelerated Genomics with Clara Parabricks, Gary Burnett
NVIDIA Clara Parabricks is a software suite for performing secondary analyzes of Next Generation Sequencing (NGS) DNA and RNA data. A big advantage of parabricks is that they are designed to deliver results at lightning fast speeds and low costs. Parabricks can analyze entire human genomes in less than 30 minutes, compared to about 30 hours for 30x WGS data. In this session, participants take a guided tour of the Parabricks suite with live examples and real applications.

Clara AGX

On-Demand: Using Ethernet to stream medical sensor data with high throughput and low latency
Medical sensors in various medical devices generate high throughput data. System designers are faced with the challenge of transferring the sensor data to the GPU for processing. In the past decade, Ethernet speeds have increased from 10G to 100G, opening up new ways to meet this challenge. We’ll explore three technologies from NVIDIA that make streaming high-throughput medical sensor data over Ethernet easy and efficient – NVIDIA Networking ConnectX NICs, Rivermax SDK with GPUDirect, and Clara AGX. Learn about the capabilities of each of these technologies and explore examples of how these technologies can be used by different types of medical devices. Finally, a step-by-step demo guides participants through installing the software, initializing a link, and testing throughput and CPU overhead.

Recommended hardware

NVIDIA Clara AGX developer kit
Provides real-time AI and imaging for medical devices. Combining low-power NVIDIA Jetson AGX Xavier and RTX graphics processors with the NVIDIA Clara AGX SDK and NVIDIA EGX stack makes it easy to securely deploy and remotely manage fleets of distributed medical instruments.
Learn more>

Discover the power of AI and robotics with the NVIDIA Jetson Nano 2GB Developer Kit. It’s small, powerful, and affordable for everyone.
Learning by doing is key for anyone new to AI and robotics, and this developer kit is great for hands-on projects.
Learn more>

NVIDIA DGX Station A100
Data science teams need a dedicated AI resource that is not available to other areas within their organization: a purpose-built, uncompromising AI system that can handle whatever busy data scientists can do, an accelerated AI platform fully optimized through hardware and software for maximum performance.
Learn more>

Kick-off spotlight

NEW to NGC: Simplify and Unify Biomedical Analytics with Vyasa
Learn how Vyasa Analytics uses Clara Discovery, Triton Inference Server, RAPIDS and DGX to develop solutions for pharmaceutical and biotechnology companies. Vyasa Analytics solutions are available in the NVIDIA NGC catalog for quick evaluation and deployment.

Do you have a start-up? Become part of NVIDIA Inception’s global network of over 8,000 startups.