Nvidia Corporation · NASDAQ:NVDA
Nvidia — Investor Day (October 2020)
34 slides from Nvidia's investor day presentation, Oct 5, 2020
Slide 1: GTC Fall 2020 Financial Analyst Event
GTC FALL 2020 FINANCIAL ANALYST EVENT
Slide 2: Forward Looking Statements
Forward Looking Statements Except for the historical information contained herein, certain matters in this presentation including, but not limited to, statements as to: our expanding ecosystem and the estimated number of AI and autonomous systems developers adopting NVIDIA AI; artificial intelligen
Slide 3: Here With You Today / Agenda
HERE WITH YOU TODAY AGENDA Computing for the Age of AI NVIDIA AI NVIDIA Enterprise AI Jensen Huang Colette Kress Founder & CEO EVP & CFO NVIDIA Edge AI NVIDIA Healthcare GTC Fall 2020 Paresh Kharya Manuvir Das Justin Boitano Kimberly Powell AI Computing Enterprise Computing Edge Computing Healthcare
Slide 4: Computing for the Age of AI
COMPUTING FOR THE AGE OF Al NVIDIA RTX NVIDIA HPC NVIDIA Al NVIDIA Enterprise Al NVIDIA Edge Al arm Mellanox TECHNOLOGIES
Slide 5: Expanding NVIDIA Ecosystem
EXPANDING NVIDIA ECOSYSTEM AI and Autonomous Systems Developers Adopt NVIDIA AI 80 2.3M New SDKs Developers 6M DLSS 2.1 CUDA Downloads in 2020; RTX DI 20M downloads to date OptiX 7.2 AERIAL HPC SDK 20.9 RTX HPC RAPIDS AI CLARA METRO DRIVE ISAAC 5G 1,800 RAPIDS 0.16 Parabricks 3.5 DeepStream 5.0 GPU-
Slide 6: NVIDIA AI Inference Leadership
NVIDIA AI INFERENCE LEADERSHIP Paresh Kharya
Slide 7: Artificial Intelligence — The Most Powerful Technology Force of Our Time
ARTIFICIAL INTELLIGENCE — THE MOST POWERFUL TECHNOLOGY FORCE OF OUR TIME EXPLODING MODEL COMPLEXITY 30,000X in 5 Years | Now Doubling Every 2 Months 1.E+04 GPT-3 HUMAN-LEVEL HUMAN-LIKE DRONE 1.E+03 READING COMPREHENSION CHATBOTS DELIVERIES Megatron-BERT 24% Better than Avg Human Half of Users Prefer
Slide 8: A100 Takes Off
A100 TAKES OFF Fastest GPU Ramp Ever May ‘20 June ‘20 July ‘20 Aug ‘20 Sept ‘20
Slide 9: 3 Pillars of NVIDIA AI
3 PILLARS OF NVIDIA AI DATA ANALYTICS & TRAINING INFERENCE APPLICATION FRAMEWORKS PRE-TRAINED MODELS TRITON SOTA Models Inference Serving Conversational AI Computer Vision Query Frameworks TIS Recommendations Result Reinforcement Learning TRANSFER LEARNING TOOLKIT TensorRT 7.2 Zero Coding | Speech,
Slide 10: Accelerating Adoption of NVIDIA AI Inference
ACCELERATING ADOPTION OF NVIDIA AI INFERENCE Celebrating 10 Years of NVIDIA GPU in the Cloud 7TH GENERATION TensorRT ACCELERATED ADOPTION TOTAL CLOUD AI INFERENCE COMPUTE CAPACITY (Inference Compiler) OF NVIDIA GPU (PETA OPS) (EXA OPS) 166 1,000,000 2000+ Optimizations 100,000 TensorFlow, PyTorch, O
Slide 11: NVIDIA AI Inference Customers
NVIDIA AI INFERENCE CUSTOMERS MICROSOFT OFFICE AMERICAN EXPRESS TENCENT PCG GRAMMAR CORRECTION FRAUD PREVENTION RECOMMENDATIONS Half a Trillion Queries per Year Over $1 Trillion in Transactions | 115M Credit Cards Videos, News, Music, Apps 200ms Inference Response 2ms Latency Enables Real-Time Fraud
Slide 12: NVIDIA AI Inference Customers
NVIDIA AI INFERENCE CUSTOMERS AUTO CONSUMER INTERNET CSP ROBOTICS MEDICAL RETAIL & FSI INDUSTRIAL
Slide 13: NVIDIA Enterprise AI: Data Center Is the New Unit of Computing
NVIDIA ENTERPRISE AI DATA CENTER IS THE NEW UNIT OF COMPUTING Manuvir Das
Slide 14: Modern Data Centers Are Software-Defined
MODERN DATA CENTERS ARE SOFTWARE-DEFINED Moving from Fixed-Function Appliances to Software Services on Servers Manual Infrastructure Management CPU VMs Containers Security Appliances Infrastructure Management Software-defined Security Software-defined Storage Software-defined Networking Storage Syst
Slide 15: Data Center Workload Growing Exponentially
DATA CENTER WORKLOAD GROWING EXPONENTIALLY Microservices CPU VMs Containers East-West Traffic Infrastructure Management Software-defined Security Software-defined Storage Software-defined Networking Storage Access NVIDIA NIC Acceleration Engines Zero Trust Security
Slide 16: Introducing the Data Processing Unit (DPU)
INTRODUCING THE DATA PROCESSING UNIT (DPU) Software-defined Data Center Infrastructure-on-a-Chip TRADITIONAL SERVER DPU-ACCELERATED SERVER CPU CPU VMs Containers Infrastructure Management Software-defined Security Software-defined Storage Software-defined Networking VMs Containers NVIDIA NIC NVIDIA
Slide 17: NVIDIA BlueField-2 DPU
NVIDIA BLUEFIELD-2 DPU Data Center Infrastructure-on-a-Chip Combines 8 64-bit A72 Arm Cores, 2 VLIW Acceleration Engines, and Mellanox Out-of-Band Management ConnectX-6 Dx NIC Accelerated Security: Isolation, Root of Trust, Crypto, Key Management, Regular Expression Engine Accelerated Networking: RD
Slide 18: NVIDIA DOCA
INFRASTRUCTURE APPLICATIONS Infrastructure Software-defined Software-defined Storage Software-defined Security Management Networking NVIDIA DOCA Data Center Infrastructure-on-a-Chip Architecture DOCA SDK for BlueField DPUs Storage Security Networking SPDK / Kernel DPDK / Kernel P4 / DPDK / Kernel Op
Slide 19: NVIDIA DPU Roadmap
1000X BlueField-4 1000 SPECINT 400 TOPS 400 Gbps NVIDIA DPU ROADMAP 100X Programmable Data Center BlueField-3X BlueField-2X 350 SPECINT Infrastructure-on-a-Chip 70 SPECINT 75 TOPS AI 60 TOPS AI 400 Gbps 200 Gbps 10X BlueField-3 350 SPECINT 1.5 TOPS 400 Gbps BlueField-2 70 SPECINT 1X 0.7 TOPS 200 Gbp
Slide 20: NVIDIA Accelerated Computing on VMware Cloud Foundation
NVIDIA ACCELERATED COMPUTING ON VMWARE CLOUD FOUNDATION Bringing AI to Every Enterprise, Enabling Next-Gen Hybrid Cloud
Slide 21: NVIDIA Edge AI: A Trillion Things – The Next Internet
NVIDIA EDGE AI A TRILLION THINGS – THE NEXT INTERNET Justin Boitano
Slide 22: The Age of AI — A Trillion Connected Things
THE AGE OF AI — A TRILLION CONNECTED THINGS 2M 13M 570M Factories Stores Farms 5M 40M 10T Call Center Miles of Road Miles Driven Agents per Year Edge Data Centers Millions of Edge DC Each with Thousands of Nodes
Slide 23: Announcing EGX Edge AI Platform
ANNOUNCING EGX EDGE AI PLATFORM NVIDIA BlueField-2 NVIDIA Ampere Architecture GPU 8x Arm Cores 3rd generation Tensor Core Dual 100 Gb/s Ethernet or InfiniBand New Security Engine for Confidential AI Line-speed TLS/IPSec Crypto Engine Secure, Authenticated Boot Time Triggered Transmission Tech for Te
Slide 24: EGX Edge AI Platform
EGX EDGE Al PLATFORM Miter CJecotonaos
Slide 25: EGX Edge AI Platform
EGX EDGE Al PLATFORM a rn oy 2 3? Isaac Ret 2 bi gpoult Y METROPOLIS °o Qc con ericsson 2 | oewanic Viasat vm
Slide 26: NVIDIA Clara Computational Drug Discovery
NVIDIA CLARA COMPUTATIONAL DRUG DISCOVERY Kimberly Powell
Slide 27: NVIDIA Clara Discovery
NVIDIA CLARA DISCOVERY $1.25T Pharmaceutical Industry | $2B R&D Per Drug | 12 ½ Years To Market | 90% Failure Rate SEARCH O(1060) Chemical Compounds RAPIDS O(109) GENOMICS STRUCTURE DOCKING SIMULATION IMAGING Biological Drug Target O(102) Candidate Schrodinger Clara Parabricks CryoSPARC, Relion Auto
Slide 28: Cambridge-1 AI Supercomputer
CAMBRIDGE-1 AI SUPERCOMPUTER 80 NVIDIA DGX A100 | 400 PETAFLOPS AI COMPUTE | 20 TERABYTES/SEC INFINIBAND | 2PB NVME | 500KW HEALTHCARE & LARGE-SCALE AI & COLLABORATION WITH UK INDUSTRY, EDUCATE FUTURE LIFE SCIENCE RESEARCH DATA SCIENCE RESEARCH UNIVERSITIES, STARTUPS AI PRACTITIONERS
Slide 29: GSK Partners with NVIDIA to Build AI Drug Discovery Lab
GSK PARTNERS WITH NVIDIA TO BUILD AI DRUG DISCOVERY LAB
Slide 30: AI Moment in Healthcare
AI MOMENT IN HEALTHCARE Accelerated Computing | Data Science & Machine Learning | Deep Learning | NLP Language Models 30X 30X BIOMEDICAL NLP Named Entity 20X Recognition 15X Vocabulary BioMegatron Relation Extraction e.g. COVID 345M Param 7X 6.1B Words 6X 6X Q&A PubMeb Genomics Structure Screening D
Slide 31: GTC Fall 2020 – Colette Kress
&* NVIDIA GTC Fall 2020 Colette Kress
Slide 32: GTC Fall 2020
GTC FALL 2020 NVIDIA AI GAINING MOMENTUM 20M CUDA downloads/6M in last year | Most shipped AI inference cloud compute capacity in the world NVIDIA POWERS ENTERPRISE COMPUTING NVIDIA AI integrated into VMware & Cloudera | NVIDIA DPUs “Data Center on a Chip Architecture" NVIDIA EDGE AI FOR THE INTERNE
Slide 33: Computing for the Age of AI
COMPUTING FOR THE AGE OF AI NVIDIA RTX NVIDIA HPC NVIDIA AI NVIDIA ENTERPRISE AI NVIDIA EDGE AI $8B $10B $45B $30B $15B 2020 Revenue* Data Center Total Addressable Market by 2024 18% Trailing 5-year CAGR *Trailing 4-quarters through F3Q 2021 outlook for Gaming and Professional Visualization market p
Slide 34: Q&A
Q&A Jensen Huang & Colette Kress