Hardware in Space
2 likes278 views
This document provides a summary of announcements and information from NVIDIA regarding their AI products and technologies. It highlights the release of the NVIDIA A100 80GB GPU, the Selene supercomputer featuring DGX A100 systems, comparisons of AI data center configurations with DGX-1 versus DGX A100 systems, and expansions to the NGC container environment and RAPIDS open source libraries. Brief descriptions of applications like SIMNET and Fourier Neural Operators are also included.
Hardware in Space
- 1. ALISON B LOWNDES AI DevRel | EMEA @alisonblowndes Happy December 2020 🎄
- 2. 2 Congrats to the Chang’e5 team! China National Space Agency
- 3. 3 ANNOUNCING NVIDIA A100 80GB Supercharging The World’s Highest Performing AI Supercomputing GPU 80GB HBM2e For largest datasets and models 2TB/s + World’s highest memory bandwidth to feed the world’s fastest GPU Multi-Instance GPU 3rd Gen NVLink 3rd Gen Tensor Core
- 4. 4 NVIDIA SELENE Now Featuring NVIDIA DGX A100 640GB 4,480 A100 GPUs 560 DGX A100 system 850 Mellanox 200G HDR switches 14 PB of high-performance storage 2.8 EFLOPS of AI peak performance 63 PFLOPS HPL @ 24GF/W
- 5. 5 TODAY’S AI DATA CENTRE 50 DGX-1 systems for AI training 600 CPU systems for AI inference $11M 25 racks 630 kW
- 6. 6 5 DGX A100 systems for AI training and inference $1M 1 rack 28 kW 1/10th COST 1/20th POWER $1M 28 kW DGX A100 DATA CENTRE
- 7. 7 NVIDIA CUDA-X AI ECOSYSTEM
- 8. 88 EXPANDING NGC NEW CONTAINERS FOR A100 & ARM Now NGC-READY SYSTEMS FOR A100 Starting Q3 NGC Private Registry NGC Container Environment Modules Higher HPC app performance w/ NVTAGS NEW FEATURES Now Multi-arch support for x86, Arm and Power Learn More – ngc.nvidia.com | NGC Private Registry | NVTAGS | NGC Container Environment Modules HPC Simulation & Visualization AI Frameworks (A100) ChromaAutoDock 4 VMD ** * Available week of June 22 ** Available starting with v20.06 * * *
- 9. 9 GPU Accelerated Data Science RAPIDS is a set of open source software libraries which gives you the freedom to execute end-to-end data science and analytics pipelines entirely on GPUs. www.rapids.ai
- 10. 10 SIMNET v.0.2 Solve larger problems faster with XLA and AMP support, and Multi-GPU, Multi- Node implementation Models Multiple Physics in Forward, Inverse and Data Assimilation simulations with accuracy & convergence Parameterized system representation to solve multiple scenarios simultaneously APIs for implementing new Physics, Geometry, and Domains and detailed User Guide examples AI-accelerated Physics Simulation Toolkit
- 12. 12 Embedded Computing without Compromise Confidential A176 - AI Fanless Small FF Supercomputer A176 – Used for: • Video processing • Recording data The ball is a recorder that is attached to A176 via a cable. There is an explosive device that will break the link with the A176, so the ball will fall to the earth without being destroyed
- 13. 13 NVIDIA AGX Family of Systems for Embedded AI HPC Self-driving cars Robotics Smart Cities Healthcare NVDLA.org
- 14. Orbital Sidekick
- 19. 19 Introducing EGX A100 Converged Accelerator Combining Mellanox and NVIDIA Ampere GPU Architecture Unprecedented AI Inference Performance In-Line Network Acceleration ●Secure GPU enclave protects AI model ●Line-speed TLS & IPSec Crypto Engines ●Service Mesh Offloads (SDN) Enhanced Security ●Ampere based Architecture ●3rd Generation Tensor Core ●Dual 100Gb/s Ethernet or InfiniBand ●Accelerated Switch & Packet Processing ●Time Triggered transmission tech for Telco (5T for 5G)
- 20. 20
- 23. 23 EO4SDG.ORG
- 24. 24
- 25. 25 INGESTION STORAGE PROCESSING SERVING BIG DATA PIPELINE Ingredients: • Lots of data • Lots of compute • Software tools • Time and patience Method: 1. Collect raw, massive sets of data. 2. Put the data in a Data Lake. 3. Grab the data that you need and sort through. 4. Find patterns in the data. 5. Solve the problem. 1. Obtaining and importing data 2. Organizing & storing data for future use 3. Manipulating and analyzing the data 4. Operationalizing the solution
- 26. 2626
- 27. 27 PHYSICS IN USD + KIT
- 29. 29 Detection Planning Acceleration Assimilation Enhancement Parametrization AugmentationPrediction Monitor Environmental Change drought flooding deforestation urbanification melting glaciers sea-level change
- 30. 30 Detection Planning Acceleration Assimilation Enhancement Parametrization AugmentationPrediction Use Inpainting to Repair Damaged GOES-17 Observations
- 32. 32
- 33. 33
- 34. Real-Time Photorealistic Digital Humans with NVIDA RTX A6000 “ Digital Domain leads the industry in pioneering real-time photo-realistic digital humans. Working with Epic Games and NVIDIA, we have continuously pushed the limits of technology. The new RTX A6000 lets us completely redefine what’s possible with real-time ray tracing and machine learning. ” Darren Hendler | Director, Digital Human Group
- 35. 35 GALACTIC WINDS NVIDIA IndeX and Cholla Brant Robertson Evan Schneider
- 36. ISAAC nvidia.com/en-gb/deep-learning-ai/industries/robotics Simulated samples of a dolly (with actual CAD model) used to train object detection and pose estimation neural networks Procedurally generated simulated images used for segmentation network training Multiple Carter robots operating simultaneously in virtual warehouse; Each operated by an independent Jetson Xavier Multi Robot HIL Simulation ML Training in Simulation Manipulation in Isaac Sim Domain Randomization (Supervised Learning) Factory of the Future Domain Randomization (Reinforcement Learning)
- 37. MUCH MORE WITH ISAAC SOFTWARE GPU Accelerated Algorithms/DNNs (GEMs) And more… Free Space Segmentation 3D Object Pose Estimation Motion Planning Stereo Depth Stereo Visual Inertial Odometry Super Pixels April Tags 2D Skeleton Pose Estimation DeepStream Integration Planner with Costmaps Multi Lidar Support Navigation (LQR Path Planner) Sensors Robot Platforms Audio
- 38. ROBOTICS MONTAGE
- 39. 39 RICH CONTENT PORTFOLIO Fundamentals and advanced hands-on training in key technologies and application domains AI for Digital Content Creation Deep Learning Fundamentals AI for HealthcareAI for Autonomous Vehicles AI for Intelligent Video Analytics Accelerated Computing Fundamentals AI for Robotics AI for Predictive Maintenance Accelerated Data Science Fundamentals Intro to AI in the Data Center AI for Anomaly Detection AI for Industrial Inspection
- 40. NVIDIA’s GTC brings together a global community of developers, researchers, engineers, and innovators with the common goal of sharing achievements while discovering new technologies and tools that drive change around the globe. If you work with any of our GPUs, DPUs, or software offerings is making a difference, submit a talk or poster to join us online in March. SHARE YOUR LIFE’S WORK AT GTC 2021 JOIN A GLOBAL COMMUNITY OF BRILLIANT MINDS ONLINE NEXT MARCH March 15 — 25, 2021 Submit your ideas at www.nvidia.com/gtc