NVIDIA AI Delivers Major Advances in Speech, Recommender Systems and Hyperscale Inference

Amazon, Microsoft, Snap, NTT Communications Deploy NVIDIA AI; NVIDIA Riva and Merlin Enter General Availability; NVIDIA AI Enterprise 2.0 Supports All Major Data Center and Cloud Platforms; NVIDIA AI Accelerated Program Launches

NVIDIA announced major updates to its NVIDIA AI platform, a suite of software for advancing such workloads as speech, recommender system, hyperscale inference and more, which has been adopted by global industry leaders such as Amazon, Microsoft, Snap and NTT Communications.

The company also announced the NVIDIA^® AI Accelerated program, which helps to ensure performance and reliability of AI applications developed by NVIDIA’s software and solution partners. The program increases visibility to a wide range of proven AI-accelerated applications, enabling enterprise customers to deploy with confidence on the NVIDIA AI platform. Adobe, Red Hat and VMware are among the more than 100 partners participating at launch.

Recommended AI News: Industry-First Event to Explore Global MPC Adoption for Digital Asset Security

“NVIDIA AI is the software toolbox of the world’s AI community — from AI researchers and data scientists, to data and machine learning operations teams,” said Jensen Huang, founder and CEO of NVIDIA. “Our GTC 2022 release is massive. Whether it’s creating more engaging chatbots and virtual assistants, building smarter recommendations to help consumers make better purchasing decisions, or orchestrating AI services at the largest scales, your superpowered gem is in NVIDIA AI.”

Freely available for developers, NVIDIA AI includes NVIDIA Riva for speech AI and NVIDIA Merlin™ for smart recommendations, now both generally available. Updates have also been made across the software suite, including tools such as the NVIDIA Triton, NeMo, Maxine and TAO Toolkit.

Additionally, NVIDIA AI Enterprise 2.0 is now optimized, certified and supported across every major data center and cloud platform, including bare-metal servers, virtualized infrastructure and CPU-only systems. The suite is now supported with Red Hat OpenShift and VMware vSphere with Tanzu.

Software Tools to Build Industry-Leading AI Applications

NVIDIA AI is comprised of key enabling SDKs and tools for rapid deployment, management and scaling of AI workloads across multiple nodes to power complex training and machine learning workloads. These include:

• NVIDIA Triton™ – NVIDIA Triton is a versatile open-source hyperscale model inference solution. The latest release contains three key updates: a Model Navigator for accelerated deployment of optimized models, Management Service for efficient scaling in Kubernetes, and Forest Inference Library enabling inference on tree-based models with explainability for fast, optimized and scalable AI in every application.
• NVIDIA Riva 2.0 – A world-class speech AI SDK that includes models pretrained with leading recognition rates, enabling developers to customize real-time speech AI applications for their industry with 2x better accuracy over generic services. Riva 2.0 includes speech recognition in seven languages, human-like deep learning-based text-to-speech with both male and female voices, as well as custom tuning with NVIDIA TAO Toolkit. NVIDIA also announced NVIDIA Riva Enterprise, a paid offering that includes enterprise support from NVIDIA.
• NVIDIA NeMo Megatron 0.9 – A framework for training large language models (LLMs), NeMo Megatron enables researchers and enterprises to train any model to convergence and scale to trillions of parameters for applications such as conversational AI, recommenders and genomics. The latest version includes new optimizations and recipes that shorten end-to-end development and training time, and adds support for training in the cloud. Early users of LLMs on NVIDIA accelerated computing include JD.com, AI Sweden, Naver and the University of Florida.
• NVIDIA Merlin 1.0 – An accelerated, end-to-end recommender AI framework to build high-performing recommenders at scale, which includes two new libraries: Merlin Models and Merlin Systems. These allow data scientists and machine learning engineers to determine which features and models are the best fit for their use case and deploy recommender pipelines as microservices.
• NVIDIA Maxine – An audio and video quality enhancement SDK that reinvents real-time communications with AI, and introduces acoustic echo cancellation and audio super resolution. The two new features enhance audio quality leading to a clearer communication experience.

Recommended AI News: Qrvey Ranked Among Top Embedded Business Intelligence Vendors by Dresner Advisory Services

Customers Embrace NVIDIA AI

Industry leaders are using NVIDIA AI to improve cost-efficiency, create more engaging customer experiences and optimize AI application capabilities.

“On Snapchat, our community plays with Lenses over 6 billion times per day,” said Alan Bekker, head of Conversational AI at Snap. “Snap is using NVIDIA Riva to optimize our AI-based speech capabilities and offer them to Lens Studio creators to build a new generation of compelling AR experiences.”

“The document translation feature within Translator, a Microsoft Azure Cognitive Service, enables efficient business to customer interactions by translating documents retaining format and structure as in source document,” said Xuedong Huang, Microsoft Technical Fellow and Azure AI CTO. “Using NVIDIA Triton, we’re able to deploy the latest Z-Code model to achieve significantly improved document translation quality with the low latency, providing our users with unmatched translation services.”

NVIDIA AI Enterprise Supports Containerized AI Across Data Centers and Cloud

An end-to-end, cloud-native suite of AI and data analytics tools and frameworks, the NVIDIA’s AI Enterprise 2.0 software suite accelerates AI development and deployment for industries. Certification for Red Hat OpenShift, the industry’s leading enterprise Kubernetes platform, enables customers to use containerized machine learning tools to more easily build, scale and share their models on bare-metal, or virtualized systems, with VMware vSphere.

“The certification of Red Hat OpenShift for NVIDIA’s AI Enterprise, and the availability of OpenShift on NVIDIA LaunchPad, unites top tools for AI development with a consistent hybrid cloud foundation,” said Stefanie Chiras, senior vice president of Partner Ecosystem Success at Red Hat. “Now, IT teams and data scientists can build and manage NVIDIA’s AI on Red Hat OpenShift, helping enterprises accelerate the delivery of intelligent applications in production.”

NVIDIA’s AI Enterprise 2.0 also introduces more NVIDIA’s AI software containers to support training and inference. Support for the NVIDIA TAO Toolkit allows enterprise developers to fine-tune and optimize NVIDIA pretrained AI models, simplifying the creation of custom, production-ready models, without AI expertise or large training data. The latest release of NVIDIA Triton Inference Server is also included in the software suite.

NTT Communications, the leading global provider of information and communications technology solutions within NTT Group, has adopted NVIDIA’s AI Enterprise to accelerate its research and development teams building NLP and intelligent video analytics applications.

“Many of our application developers now use accelerated computing, and are in need of an internal infrastructure that provides an easy-to-use, cost-effective GPU-enabled environment,” said Shoichiro Henmi, director of Technology Division, Innovation Center, NTT Communications. “We are confident that NVIDIA’s AI Enterprise will provide an ideal solution as an AI-enabled platform to support large-scale development in our VMware vSphere, Kubernetes and cloud infrastructure.”

Recommended AI News: Solodev Launches Serverless Health Cloud to Power Fast, Secure, and Scalable Patient Portals on AWS

[To share your insights with us, please write to sghosh@martechseries.com]