Talk on commercialising space data
0 likes369 views
This document discusses NVIDIA's efforts to move AI and accelerated computing technologies from research applications to real-world deployments across various domains. It outlines NVIDIA's hardware and software stack including GPUs, DPUs, CPUs and frameworks that can rearchitect data centers for AI. It also highlights several application areas like climate science, drug discovery, cybersecurity where NVIDIA is working to apply AI at scale using technologies like accelerated computing and graph neural networks.
Talk on commercialising space data
- 1. MOVING FROM RESEARCH TO DEPLOYMENT ACROSS SPACE SCIENCE ALISON B LOWNDES | SENIOR SCIENTIST | GLOBAL AI
- 4. 4 10.5 MW 45,000 Sq. Ft.
- 5. 5 NVIDIA SELENE Now Featuring NVIDIA DGX A100 640GB #1 in Green 500 | #5 Top500 | #1 MLPerf | #1 Industrial System 4,480 A100 GPUs 560 DGX A100 640GB systems 850 Mellanox 200G HDR switches 14PB of high-performance storage 2.8 EFLOPS of AI peak performance 63 PFLOPS HPL @ 24GF/W 1 DGX A100 replaces 150 CPU servers Saves 300 tons of CO2 per DGX per year!
- 6. 6 MOTLEY FOOL “NVIDIA IS NOT JUST A GRAPHICS CHIP COMPANY ANYMORE” GPU Ampere A100 CPU Grace DPU BlueField-3 Next-generation data centers are an orchestration of three pillars: the GPU for accelerated computing, the CPU for general-purpose computing, and the DPU, which processes and moves data in the data center. The introduction of the NVIDIA Grace™ CPU and the NVIDIA BlueField® DPU make NVIDIA a three-chip company, aimed at rearchitecting the data center for AI.
- 7. NVIDIA Accelerated Computing Full Stack, 3 Chips, Data Center Scale 30 Million CUDA Downloads 150 SDKs $100 Trillion Industry Served Gaming Data Science Robotics Broadcast CAD Physical Sciences Life Sciences Quantum Physics Digital Twins Genomics 5G Quantum Computing Cybersecurity AI NLU Machine Learning AI Recsys AI Speech AI Computer Vision Medical Imaging Autonomou s Vehicles Edge
- 9. 9 BUILDING AN ENTERPRISE AI PLATFORM AI Infrastructure AI Application Development Data Prep Training Inference Accelerated Computing High-Performance Storage and Networking Prototype IT infrastructure optimized for AI workloads speeds model development
- 10. AI INFERENCE IS HARD PROCESSORS AI Inference DEPLOYMENT PLATFORMS Cloud On-Prem Edge Embedded T4 GPU Arm CPU A100 GPU V100 GPU x86 CPU FRAMEWORKS APP CONSTRAINTS Real Time Batch Streaming MODELS CNN GNN Decision Trees RNN Transformers
- 12. 12 12 Pretrained Models TLT Customize & Speed Up Model Training Trained Models TensorRT Optimize For Multiple Constraints For High Perf. Inference On GPU Triton AI Application Query Result Proprietary Data Model Store INFERENCE WITH NVIDIA AI ON GPUS Scaled Multi-framework Inference Serving For High Perf. & Utilization On GPU/CPU
- 13. NGC CONTAINERS ENABLE YOU TO FOCUS ON BUILDING AI PERFORMANCE OPTIMIZED DEPLOY ANYWHERE ENTERPRISE READY SOFTWARE Scanned for CVEs, malware, crypto Tested for reliability Backed by Enterprise support Scalable Updated monthly Better performance on the same system Docker | cri-o | containerd | Singularity Bare metal, VMs, Kubernetes Multi-cloud, on-prem, hybrid, edge https://catalog.ngc.nvidia.com/resources
- 14. 37 TAO TOOLKIT WORKFLOW Automatic Mixed Precision | Quantization Aware Training | Pruning https://developer.nvidia.com/tao-toolkit
- 15. 36
- 16. On-premises In the cloud https://github.com/rapidsai Source code on GitHub Containers on NGC & Docker Hub https://ngc.nvidia.com Conda packages https://anaconda.org/rapidsai Pascal architecture or better CUDA 9.2, 10.0 or 10.1.2 Ubuntu 16.04/18.04, CentOS 7 & RHEL 7
- 18. FDL 2022 Predicting solar winds & storms for insurance risk mitigation Assessing solar drag effects on satellites & orbital control Super-res for satellite imagery across all computer vision applications Improving extreme weather prediction & wildfire / flood detection Digital twinning of coastlines – as well as full Earth (E-2) Automatic labelling for events like wildfire or oil spills (UN WFP) Assessing solar drag effects on satellites & orbital control On-board processing, in orbit. Oceanic carbon sequestration & marine restoration Data fusion & burn rate prediction in bushfires (NSW Govt)
- 19. 16 “NVIDIA is proud to have been one of ‘the Mavericks’ at the inception of the Frontier Development Laboratory. Our headquarters are named Endeavor and Voyager – mighty ships for our journey to the stars. We are delighted to support FDL as we push the frontiers for the human race.” JEN-HSUNHUANG PRESIDENT & CEO, NVIDIA CORPORATION
- 21. fll:(INTl II CE"l;LOP'Ml;NT L.AIIJI Gesa Google Clloud - ---- ---- --= -= - - -- ==-== =· AIRBUS @ n e t . IWMI - - - - C A T A P U L T --. -- UUJVERSITV OF OXFORD T R IL LIU M
- 23. LUNAR SURFACE IMAGERY ENHANCEMENT 20 Fig. 1: NASA’s Lunar Reconnaissance Orbiter Fig. 2: Apollo XVI landing site
- 24. Next Level of AI GPGPU in Space Applications Aitech’s S-A1760 Venus™: most powerful and smallest space AI GPGPU in small form factor (SFF). Suitable for the next gen of short duration spaceflight, NEO and LEO. 22
- 25. Orbital Sidekick
- 27. How do we understand our world?
- 30. MILLION-X CLIMATE SCIENCE 1000km 100km 10km 1km 100m 10m 1m 1980 1990 2000 2010 2020 2030 2040 2050 2060 AR2 AR3 AR4 AR5 AR6 (IPCC) AR1 1km at 1min (1X COMPUTE) 100m at 1s (10,000X COMPUTE) 1m at 0.01s (100 BILLION X COMPUTE) CONVECTION RESOLVING STORM RESOLVING STRATOCUMULU S RESOLVING RESOLUTION Figure adapted from: Schneider, T., Teixeira, J., Bretherton, C. et al. Climate goals and computing the future of clouds. Nature Clim Change 7, 3–5 (2017). https://doi.org/10.1038/nclimate3190 MILLION-X CLIMATE SCIENCE
- 31. EARTH DIGITAL TWIN IN OMNIVERSE ERA5 ECMWF Atmospheric Winds & Geopotential 10 TB | 30km | 5 Atmos Layers 100,000X Speed-Up 0.25 Seconds for 7-Day Forecast Training:4 Hours on 128 A100 GPUs RAPIDS Modulus Omniverse Fourier Neural Operator AI EARTH EMULATOR Satellite Ocean Ecosystem Atmosphere Extreme Weather Prediction Wind Energy Forecasting Disaster Mitigation
- 32. GRAPH NEURAL NETWORKS CAPTURE INSIGHTS FROM INTERCONNECTED DATA 90B Relationships in a Social Network 1.1B Transactions a Day 40B Molecules in Chemical Databases DRUG DISCOVERY SOCIAL CONNECTIONS | FRAUD DETECTION
- 33. ANNOUNCING DGL ACCELERATION WITH CUDA-X GPU-Accelerated GNN Workflow GNN for Molecule Reaction Prediction, Node Classification, Knowledge Graphs, and Model Explainability CUDA-Optimized Reference Examples for SE3-Transformer, R-GCN, and GraphSage Early Access in December ngc.nvidia.com cuDF cuGraph DGL with CUDA- X Text Images Relationships Sub-Graph Construction Graph to GNN Graph Construction
- 34. LEVERAGE AI INFERENCE INSTEAD OF HEURISTICS Outdated Rules Engines Cannot Recognize Patterns of Emerging Cyber Threats PERCEIVE INFER ACT
- 35. NVIDIA MORPHEUS CYBERSECURITY FRAMEWORK A Platform to Defend Against Tomorrow’s Threats SENSOR TELEMETRY COMPUTE PLATFORM PERCEIVE ACT INFER
- 36. NVIDIA MORPHEUS CYBERSECURITY FRAMEWORK A Platform to Defend Against Tomorrow’s Threats BLUEFIELD 2 NVIDIA MORPHEUS TRITON AI Language Models PERCEIVE ACT DPU SENSOR TELEMETRY COMPUTE PLATFORM INFER GPU Server
- 37. Anomaly Detection Machine Human Machine Machine Machine Machine Human Humans and Machines Across the Enterprise Post-Processing Pre-Processing Inference Requests Apache Kafka TRITON Log Data NVIDIA MORPHEUS ANNOUNCING NVIDIA MORPHEUS Accelerated AI Platform for Next Gen SIEM Built on NVIDIA RAPIDS and NVIDIA AI 600X Faster Data Processing – Monitor Every User and Machine-Generated Data for Anomalous Behavior Detect Anomalies with 10s of Millions of AI Models in Real-Time Pre-Trained Models for User Activity Fingerprinting and Phishing Detection Early Access 2 Available Now nvidia.com/morpheus
- 41. LIVE SYNTHETIC DATA GENERATION PIPELINE
- 42. NVIDIA MODULUS A Framework for Developing Physics-ML Models for Digital Twins ▪ Train Physics-ML Models Using Governing Physics, Simulation, and Observed Data ▪ Multi-GPU Multi-Node Training ▪ 1,000-100,000X Speed Models – Ideal for Digital T wins ▪ Available now
- 43. NVIDIA CONFIDENTIAL. DO NOT DISTRIBUTE.
- 44. 44 NVIDIA A100X AND NVIDIA A30X A100 / 30 Tensor Core GPU BlueField-2 DPU ConnectX-6 Dx 8 ARM A72 Cores Integrated PCIe Gen4 Switch 2 slot FHFL 300 / 230 W Converged Accelerators AI-Based Security 5G vRAN AI on 5G
- 45. 45 NVIDIA CONVERGED ACCELERATOR DEVELOPER KIT Get Started Hardware Access A30X units for select customers and partners Documentation API and Runtime documentation Sample Application CUDA + DOCA sample applications https://developer.nvidia.com/converged-accelerator-devkit
- 49. NVIDIA-CERTIFIED SERVERS NVIDIA-CERTIFIED WORKSTATIONS POWERED BY NVIDIA RTX NVIDIA STUDIO PROFESSIONAL LAPTOPS
- 50. 50 EXPLORE OMNIVERSE ENTERPRISE SEE YOU IN OMNIVERSE GET ACCESS TO A FREE TRIAL DEVELOP ON OMNIVERSE DOCUMENTATION docs.omniverse.nvidia.com FORUMS omniverse.nvidia.com/forums TUTORIALS AND WEBINARS omniverse.nvidia.com/tutorials DISCORD discord.gg/nvidiaomniverse
- 51. 66
- 52. NVIDIA DRIVE AV Synthetic Data World Data VIRTUAL WORLD PHYSICAL WORLD HD Map TRAIN AI MODEL
- 53. TRANSFORM SURROUND 2D TO 4D WORLD MODEL
- 54. JETSON AGX ORIN Next Level AI Performance for next-gen robotics Server Class AI Performance at the Edge - 200 INT8 TOPS – Ampere Tensor Cores GPU +DLA - 12x A78 ARM CPU - Up to 64 GB memory, 204 GB/s - Devkits* will be available Q1, 2022 * The Jetson AGX Orin Developer Kit will be available with 32GB memory in Q1.
- 55. Silicon Valley Singapore Paris Washington, D.C. Dallas Amsterdam Frankfurt London Tokyo ANNOUNCING NVIDIA LAUNCHPAD ACROSS THE GLOBE
- 56. HELPING YOU SOLVE YOUR MOST CHALLENGING PROBLEMS • Build and deploy end-to-end projects across a range of technologies and domains. • Gain hands-on experience with the most widely used, industry-standard software, tools, and frameworks. • Join live, instructor-led workshops and learn from DLI-certified instructors who are experts in their fields. • Take self-paced, online courses anytime, anywhere. • Earn NVIDIA DLI certificates to demonstrate subject matter competency and support career growth. Hands-On Training in AI, Accelerated Computing, Accelerated Data Science, Graphics and Simulation, and More Learn more: www.nvidia.com/dli Deep Learning Fundamentals AI for Anomaly Detection AI for Industrial Inspection Conversational AI AI for Intelligent Video Analytics Accelerated Computing Fundamentals Recommender Systems AI for Predictive Maintenance Accelerated Data Science Fundamentals Graphics and Simulation Networking AI in the Data Center
- 57. 8 2 NVIDIA DEVELOPER PROGRAM JOIN THE COMMUNITY THAT’S CHANGING THE WORLD TOOLS • Get exclusive access to an extensive library of NVIDIA software, spanning all of NVIDIA’s technology platforms. • Save time with ready-to-run, GPU-optimized software, model scripts, and containerized apps from the NVIDIA NGC™ catalog. • Participate in early access programs where you can be one of the first to experience the latest NVIDIA technology. TRAINING • Take advantage of research papers, technical documentation, developer blogs, and industry-specific resources. • Choose from a broad catalog of training options through the NVIDIA Deep Learning Institute (DLI). • Get unlimited access to NVIDIA On-Demand, the home for NVIDIA resources from GTCs and other leading industry events. COMMUNITY • Network with like-minded developers, engage with GPU experts, and contribute to discussions in the developer forums. • Attend exclusive meetups, GPU hackathons, and events. • Connect with NVIDIA experts through developer-focused webinars and Instructor-led workshops. Join the Free Program developer.nvidia.com/join
- 58. 8 3 NVIDIA INCEPTION PROGRAM Over 10,000 AI startups and growing Access to technology expertise Go-to-Market support Venture Capital Funding & Ecosystem nvidia.com/inception
- 59. SHARE YOUR INNOVATIVE WORK AT GTC 2022. Join The World's Most Brilliant Minds Online, March 21-24, 2022 NVIDIA’s GTC brings together a global community of developers, researchers, engineers, and innovators with the goal of sharing achievements while exploring breakthroughs and tools that drive growth around the globe. This is your opportunity to present your latest work to many of the world’s brightest minds within the automotive industry and beyond. There will also be hundreds of sessions covering a variety of industries and technologies, including AI, autonomous vehicles, robotics, high- performance computing, graphics, data science, and more. www.nvidia.com/gtc March 21 – 24, 2022
- 60. THANK YOU alowndes@nvidia.com | @alisonblowndes