Nvidia Corporation Stock Price, News & Analysis

NVIDIA has introduced the NeMo Megatron framework alongside the Megatron 530B customizable large language model and Triton Inference Server to facilitate the development of advanced language-based AI applications. This framework allows enterprises to create domain-specific chatbots and AI models with substantial scalability across GPUs. The framework is designed for easier data management and efficient training of language models, which can significantly improve real-time performance for AI applications.

NVIDIA has unveiled the Quantum-2 InfiniBand networking platform featuring a 400Gbps switch designed to meet the demands of AI, data analytics, and HPC applications. This advanced networking solution doubles the speed and triples the number of ports compared to its predecessor. It enhances performance by 3x while reducing power consumption and the need for additional data center switches. The platform is equipped with features for multi-tenancy and advanced telemetry for reliable throughput amidst varying user demands. It aims to cater to growing cloud computing and supercomputing service needs.

NVIDIA will host a conference call on November 17, 2021, at 2 p.m. PT to discuss its third-quarter financial results for fiscal year 2022, which ended on October 31, 2021. Prior to the call, written commentary from the CFO will be shared, available on the investor website at approximately 1:20 p.m. PT. The live webcast will allow analysts and institutional investors to engage in a Q&A session. NVIDIA is recognized for its innovation in GPUs, reshaping industries like transportation and healthcare.

NVIDIA has unveiled its next-generation cloud gaming platform, offering GeForce RTX 3080-class performance exclusively through a new membership tier. The GeForce NOW RTX 3080 membership provides gamers with access to enhanced resolutions and frame rates, utilizing over 1,000 GPUs in its SuperPOD infrastructure, delivering more than 39 petaflops of graphics power. Memberships are priced at $99.99 for six months, with early access for existing members. Availability is set for North America in November and Europe in December. This advancement underscores NVIDIA's leadership in the gaming technology space.

NVIDIA has expanded its collaboration with Lenovo by including the latter in its early access program for Project Monterey, aimed at modernizing enterprise data centers. This initiative utilizes NVIDIA BlueField DPUs to enhance security and performance in data centers. The partnership enables enterprises to leverage VMware-enabled preconfigured clusters accelerated by BlueField-powered servers. This collaboration underscores the drive towards AI integration in modern workloads, improving data center architectures.

NVIDIA announced the expansion of its cloud gaming service, GeForce NOW, with new titles from Electronic Arts. Gamers can now enjoy Battlefield™ 1 Revolution, Mirror's Edge™ Catalyst, Unravel™ Two, Dragon Age™: Inquisition, and Apex Legends™ on underpowered devices. Membership has doubled, reaching 12 million gamers worldwide. The service allows access to popular titles without additional purchases. This expansion is vital as nearly 80% of GeForce NOW users utilize devices not capable of running the latest PC games, making gaming more accessible globally.

NVIDIA will host a global, virtual GTC conference from Nov. 8-11, 2021, featuring a keynote by CEO Jensen Huang and prominent speakers like Fei-Fei Li and Tim Sweeney. Over 200,000 participants are expected to discuss advancements in AI, deep learning, and high-performance computing. The event will also include specialized tracks for startups and educational workshops through the Deep Learning Institute. Registration is free, promoting accessibility for diverse global communities.

NVIDIA announced the deployment of the largest GPU-based supercomputer, Polaris, at the U.S. Department of Energy's Argonne National Laboratory. This supercomputer features 2,240 NVIDIA A100 Tensor Core GPUs, achieving nearly 1.4 exaflops in theoretical AI performance and 44 petaflops in double-precision performance. Designed for high-performance computing, Polaris will support advanced research projects across various fields, including cancer treatment and clean energy. Researchers from academia and industry will have access through the ALCF’s allocation programs.

NVIDIA has launched NVIDIA AI Enterprise, enabling enterprises to virtualize AI workloads on VMware vSphere and standard servers. This software suite allows companies like Cerence and the University of Pisa to harness AI capabilities for various applications. Additionally, partnerships with leading manufacturers ensure optimized solutions for AI workloads. Subscription licenses start at $2,000 per CPU socket, with options for enhanced support, marking a significant advancement in enterprise AI.