Powering the AI revolution: Oracle at NVIDIA GTC

March 1, 2024 | 6 minute read
Richard Wang
Solutions Architect - NVIDIA
Sanjay Basu
Sr. Director, AI/ML GPU Services
Text Size 100%:

The authors want to thank Faith Nishimura from NVIDIA for her valuable support in this blog publication.

Industry luminaries, developers, researchers, and business strategists are gathering online and at the San Jose Convention Center for NVIDIA GTC, the conference for the era of AI, running March 18–21, 2024. With over 900 sessions, more than 200 exhibits, over 20 technical workshops, and tons of networking events, NVIDIA GTC showcases the latest advancements in AI, high-performance computing, and more, and delivers a prime opportunity for peers to connect and knowledge-share. As a Diamond Elite sponsor of GTC, Oracle is focused on integrating GPU-accelerated computing and AI to deliver powerful solutions to its customers.

Oracle at GTC

Oracle is collaborating with NVIDIA to revolutionize the computing landscape, providing solutions that harness the power of Oracle Cloud Infrastructure (OCI) with NVIDIA accelerated computing and showcasing the following services:

  • OCI GPU HPC services, offering customers NVIDIA L40S GPU, NVIDIA A10 Tensor Core GPU, NVIDIA H100 Tensor Core GPU and A100 Tensor Core GPU instances for training and inference workloads.
  • NVIDIA AI Enterprise, an end-to-end cloud native software platform that accelerates data science pipelines and streamlines development and deployment of production-grade AI applications, including generative AI. NVIDIA AI Enterprise  provides regular security updates, assurance of API stability, workload and infrastructure manageability, and NVIDIA Enterprise Support.
  • NVIDIA DGX Cloud on OCI, an AI-training-as-a-service platform that includes NVIDIA AI Enterprise.

Oracle’s collaboration with NVIDIA enables breakthroughs in AI, machine learning (ML), and data analytics, allowing customers to run high-performance computing workloads, develop AI-driven applications, and unlock insights from massive datasets with unprecedented speed and efficiency.

Join Oracle at booth #1416 to explore the achievements of our collaboration, featuring engaging live demos and Q&A sessions with AI experts. Visit the NVIDIA Healthcare Pavilion to explore Oracle’s healthcare solutions and witness a demo on Clinical Trial Patient Cohort Generations with large language models (LLMs) and retrieval-augmented generation (RAG) in collaboration with Quantiphi.

 

Oracle sessions at GTC

Check out Oracle’s in-person and on-demand sessions:

Charting the Future of AI in MedTech [S62680]

  • Speakers: David Niewolny, director of Business Development and Healthcare/Medical at NVIDIA
    Parminder Bhatia, chief artificial intelligence officer at GE Healthcare
    Dr. Sarah Matt, vice president of product strategy at Oracle Health
    Shan Jegatheeswaran, VP of Global Head MedTech Digital at Johnson & Johnson
    Rodolphe Katra, vice president of Artificial Intelligence at Medtronic
  • Time: Tuesday, March 19, 2:00 –2:50 p.m. PT
  • Summary: Artificial intelligence is revolutionizing the way diseases are prevented, diagnosed, and treated, while reducing the healthcare system cost curve and improving access to care. Join some of the most influential leaders in the medical technology space to learn how they leverage this transformative technology in various parts of their value chain, from design and development to application and deployment of medical devices. In this session, we discuss the following concepts:
  • How AI-driven medical devices are set to play a pivotal role in delivering advanced, patient-centric care, and democratizing access to care
  • Opportunities and challenges of creating a dynamic ecosystem for software-as-medical device (SaMD) platforms
  • The potential of AI, connected devices, digital twinning, and data-driven insights, and their interplay with innovative business models in the MedTech sector

 

How OCI and NVIDIA Can Power Generative AI and LLMs (Presented by Oracle) [S63064]

  • Speaker: Sagar Rawal, vice president of Compute Product Management at Oracle
  • Time: Tuesday, March 19, 3:00–3:25 p.m. PT
  • Summary: Generative AI powered by cloud infrastructure is enabling a plethora of innovative use cases. To get here, we had both hardware and software advancements to deploy AI models at scale. Come learn about OCI AI infrastructure and GPU capabilities and workloads, such as NVIDIA AI Enterprise, NVIDIA DGX Cloud, and NVIDIA Omniverse. Get an overview of our GPU compute products for small training, inference, and graphics and large scale-out training for tens of thousands of GPUs. We explore the latest GPU products on OCI, including the NVIDIA H200 Tensor Core GPU, NVIDIA GH200 Grace Hopper Superchip, and the NVIDIA L40S GPU. You’ll also learn more about our supercluster architecture powered by ultra-low-latency remote direct memory access networking for faster AI training using massive clusters.

 

HPC Quantitative Finance and Generative AI Convergence: OCI With NVIDIA GPUs Sets Financial Risk Calculation Records (Presented by Oracle) [S63251]

  • Speakers: Xinghong He, master principal cloud architect at Oracle
    Prabhu Ramamoorthy, global partner manager of Trading and Capital Markets on the Financial Services and Technology team at NVIDIA
  • Time: Tuesday, March 19, 3:30–3:55 p.m. PT
  • Summary: Convergence of HPC and AI systems using Oracle Cloud Infrastructure plus NVIDIA accelerated computing is driving the latest in trading and capital markets for speed, higher return on investments, reduced total cost of ownership, and maximum productivity with tooling, software, and language choices at end customers.

 

Accelerating Innovation in Healthcare and Life Sciences [EXPT63236]

  • Speaker: Dan Spellman, global AI cloud director at Oracle
  • Time: Tuesday, March 19, 5:00–5:15 p.m. PT
  • Summary: Today, Oracle is positioned at the intersection of having the fastest-growing cloud (OCI) and the largest global electronic health record market share through its $28 billion acquisition of Cerner (now Oracle Health). Helping fuel this ecosystem is its unique partnership with NVIDIA. Hear from Dan Spellman, global AI cloud director for Oracle, as he explores how Oracle and NVIDIA are enabling healthcare and life science organizations to drive breakthrough innovations. From large language models to genomic analysis, Dan and other industry leaders highlight recent advancements in generative AI and what lies ahead.

 

Enable Hybrid Training and Inference With NVIDIA DGX Cloud and OCI GPU Infrastructure (Presented by Oracle) [S63030]

  • Speaker: Dr. Sanjay Basu, senior director of Oracle Gen AI & GPU Cloud Engineering at Oracle
  • Time: Wednesday, March 20, 4:00–4:25 p.m. PT
  • Summary: In this session, we present a comprehensive overview of the integration and optimization of LLM training using NVIDIA DGX Cloud on OCI, coupled with efficient inference deployment on OCI's GPU infrastructure. We specifically focus on how the NVIDIA Triton Inference Server MLOps pipeline plays a pivotal role in deploying trained models seamlessly from DGX Cloud to NVIDIA GPU shapes on OCI, thereby enabling a robust and scalable hybrid AI infrastructure.

 

Improve the Quality of Large Language Model-Generated Responses with OCI GenAI Services and Qdrant Vector Database [S63066]

  • Speaker: Dr. Sanjay Basu, senior director of Oracle Gen AI & GPU Cloud Engineering at Oracle
  • Time: On-Demand
  • Summary: Vector databases are well-suited to generative AI, especially in LLM apps with RAG. Learn how to enhance your applications with contextual data insights using OCI Generative AI services and Oracle’s semantic search capabilities across Oracle Database and Oracle OpenSearch.

 

Fine-Tune Large Language Models for Higher Performance on Your Specific Use Case to Better Serve Customers [S63031]

  • Speakers: Jon Shelley, consulting member of technical staff for AI and HPC OCI at Oracle
  • Devon Crouse, consulting member of technical staff for Container & Kubernetes Services at Oracle
  • Yang Liu, senior solutions architect at NVIDIA
  • Time: On-Demand
  • Summary: Companies want to better understand LLMs and how they fit into their overall strategy. They have many general-purpose LLMs to choose from. When they’ve selected an LLM, it's beneficial to fine-tune it to best serve their customers' needs. In this session, we use the NVIDIA Nemotron-3 8B model and show how to fine-tune it with a representative dataset on OCI NVIDIA H100 systems using an Oracle Kubernetes Engine (OKE) environment. After the model has been fine-tuned, it’s optimized with the NVIDIA TensorRT-LLM library and deployed using OCI’s NVIDIA A10 and A100 systems on OKE to show how one can scale the model up based on the customer demand.

 

We hope to see you there!

Register for GTC with our referral and get 20% off your conference pass. Come visit Oracle at booth #1416 and see all we have to offer here. See you there!

Richard Wang

Solutions Architect - NVIDIA

Richard is an accomplished Solutions Architect with over 15 years of experience in the technology industry, specializing in AI/ML and cloud technologies. He has a proven track record of designing, implementing, and managing robust, scalable solutions that drive business performance and competitive advantage.
As a thought leader in AI/ML and cloud technologies, Richard has been a speaker at numerous tech conferences and has published articles in respected tech journals. He is passionate about pushing the boundaries of what is possible and sharing his knowledge to inspire others.

Sanjay Basu

Sr. Director, AI/ML GPU Services

Sanjay focuses on OCI's NVIDIA GPU offerings for large scale model training and inference. He also works with Oracle AI, Blockchain, Microservices along with Cloud Security and Compliance.


Previous Post

Maximizing efficiency and savings: Using Avesha Smart Scaler for cost-effective Kubernetes clusters in OKE

Mayank Kakani | 4 min read

Next Post


Data migration to and between OCI storage services using Resilio Connect

Ashutosh Mate | 5 min read