2. 2
NVIDIA
FUELLING THE AI
REVOLUTION
WITH GAMING
PROJECT SOL:
A Showcase for the Power of NVIDIA RTX
MINECRAFT RTX:
Real-time Ray Tracing in the World’s Most Popular Game
OMNIVERSE:
A Powerful Collaboration Platform for 3D Design
NASA MARS LANDER:
Visualizing NASA’s Supercomputer Simulations
6. 66
25 YEARS OF ACCELERATED COMPUTING
X-FACTOR SPEED UP FULL STACK ONE ARCHITECTURESYSTEMS
GPU
CPU
7. 7
ENABLING ENTERPRISE TRANSFORMATION WITH AI
End to End Application Frameworks
Desktop Development Data Center Solutions Accelerated Edge Supercomputers GPU-Accelerated Cloud
Jarvis Merlin Metropolis Clara Isaac Drive Aerial
Conversational
AI
Recommender
Systems
Smart Cities Healthcare Robotics Autonomous
Vehicles
Telecom
10. 10
NVIDIA’s DGX SuperPOD Deployment
THE SELENE SUPERCOMPUTER
27.6 PetaFLOPS HPL
20.5 GigaFLOPS/watt
1+ ExaFLOPS AI
Built from the NVIDIA DGX SuperPOD Reference Architecture
2nd generation SuperPOD Reference Architecture
NVIDIA DGX A100 and NVIDIA Mellanox InfiniBand network
The blueprint for AI power and scale, infused with the knowhow
from NVIDIA’s decade plus of AI experience
Configuration:
2,240 NVIDIA A100 Tensor Core GPUs
280 NVIDIA DGX A100 systems
494 Mellanox 200G HDR InfiniBand switches
7 PB of all-flash storage
One of the fastest and most efficient
supercomputers on the planet —
built in under one month
nvidia.com/en-us/data-center/dgx-a100/
14. 14
NVIDIA BREAKS ALL 16 RECORDS IN AI PERFORMANCE
Among Commercially Available Systems
Benchmark Max Scale Records
(DGX SuperPOD)
Per Accelerator Records
(NVIDIA A100)
Recommendation (DLRM) 3.33 Min 0.44 Hrs
NLP (BERT) 0.81 Min 6.53 Hrs
Reinforcement Learning (MiniGo) 17.07 Min 39.96 Hrs
Translation (Non-recurrent) Transformer 0.62 Min 1.05 Hrs
Translation (Recurrent) GNMT 0.71 Min 1.04 Hrs
Object Detection (Heavy Weight) Mask R-CNN 10.46 Min 10.95 Hrs
Object Detection (Light Weight) SSD 0.82 Min 1.36 Hrs
Image Classification (ResNet-50 v1.5) 0.76 Min 5.30 Hrs
Per Chip Performance arrived at by comparing performance at the same scale when possible. Per Accelerator comparison using reported performance for MLPerf 0.7 NVIDIA A100 (8 A100s). MLPerf ID DLRM: 0.7-17, ResNet50 v1.5: 0.7-
18, 0.7-15 BERT, GNMT, Mask R-CNN, SSD, Transformer: 07-19, MiniGo: 0.7-20. Max Scale: All results from MLPerf v0.7 using NVIDIA DGX A100 (8xA100s). MLPerf ID Max Scale: ResNet50 v1.5: 0.7-37, Mask R-CNN: 0.7-28, SSD: 0.7-33,
GNMT: 0.7-34, Transformer: 0.7-30, MiniGo: 0.7-36, BERT: 0.7-38, DLRM: 0.7-17.
MLPerf name and logo are trademarks. See www.mlperf.org for more information.
15. 15
NVIDIA JARVIS
Fully Accelerated Framework for Multimodal Conversational AI Services
End-to-End Multimodal Conversational AI Services
Pre-trained SOTA models-100,000 Hours of DGX
Retrain with NeMo
Interactive Response – 215ms on V100 versus 25sec on CPU
Deploy Services with One Line of Code
RETRAIN
video
audio
Multi-Speaker
TranscriptionNVIDIA GPU CLOUD NVIDIA AI TOOLKIT
Transfer Learning
NeMo
Service Maker
TRITON INFERENCE SERVER
Dialog Manager
Chatbot
Multi-
Speaker
Transcription
Look to Talk
Gesture
Recognition
Speech
Vision
NLU
Jarvis
Sign up for Early Access
https://developer.nvidia.com/nvidia-jarvis
24. Currently get 1 day extreme rain forecast
which makes her crop vulnerable.
What if we could improve forecasting to
a 5 day forecast?
25% destroyed 95% harvested
26. 26
AI REVOLUTION
AI in Space is More Critical Than Ever
Earth Observations Urban Planning Solar PredictionsPlanetary Defense
Planetary Exploration Manufacturing LunarMission Control
$400B INDUSTRY — Increase operational efficiency and safety across every challenge using AI
27. 27
BUILDING AN AI MODEL
AI MODELFEATURES DEPLOYMENTDATA
DATA
ANALYTICS
MACHINE
LEARNING
MODEL
VALIDATION
NEW DATA
28. 28
BUILDING AN AI PRODUCT
SENSORS
PERCEIVE REASON
PLAN
DATA
DATA
ANALYTICS
MACHINE
LEARNING
AI MODEL
VALIDATION
ACTUATORSAI MODEL
32. 32
NVIDIA ROBOTICS
RESEARCH LAB SEATTLE
Drive breakthrough robotics research
to enable the next-generation of
robots that safely work alongside
humans, transforming industries such
as manufacturing, logistics,
healthcare, and more
42. 42
JETSON NANO
$99 NVIDIA CUDA-X AI Computer
CUDA-X acceleration stack | High-resolution sensor support | Runs all CUDA-X AI models
43. 43
NVIDIA JETSON
Software-Defined AI Platform
Sensor Fusion & Compute Performance Expertise, Time to Market
JETSON COMPUTER
ECOSYSTEMSOFTWARE DEFINED
Jetpack SDK ∙ CUDA ∙ TensorRT ∙ TensorFlow ∙ ONMX ∙ ROS
Artificial Intelligence Computer Vision
Accelerated Computing Multimedia
Gesture rec
Obj detectPath planningDepth est
Pose est Speech rec
SDK, Design Tools, Libs, GEMs
Act
Sense
Reason
AI at the Edge
44. 44
DEEP LEARNING INSTITUTE
Training Labs
Nanodegrees
nvidia.com/DLI
TWO DAYS TO A DEMO
Create your first demo today
developer.nvidia.com/
embedded/twodaystoademo
JETSON DEVELOPER KIT
AGX Xavier Developer Kit $699
Xavier NX software patch
developer.nvidia.com/
buy-jetson
GTC
Largest event for GPU
developers
gputechconf.com
JETSON - START NOW
45. 45
CLOUD NATIVE ON JETSON
Easier and Faster Deployment
- Eliminates complex, time-consuming builds and installs
Agile and Easier Development
- Update specific modules not the whole system
Scalability
- Push the right container to the right platform
Portable
- Deploy across various environments, from test to
production with minimal changes
Consistency Across Portfolio
- Update once and push broadly
Containers & Micro Services
46. 46
GET STARTED WITH NGC
Deploy containers:
ngc.nvidia.com
Learn more about NGC offering:
nvidia.com/ngc
Technical information:
developer.nvidia.com
Explore the NGC Registry for DL, ML & HPC
47. 47
PURPOSE BUILT PRE-TRAINED NETWORKS
Number of classes: 3
Dataset: 750k frames
Accuracy: 84%
Number of classes: 4
Dataset: 150k frames
Accuracy: 84%
Number of classes: 12
Dataset: 56k frames
Accuracy: 88%
Number of classes: 20
Dataset: 60k Frames
Accuracy: 92%
Number of Classes: 4
Dataset: 160k frames
Accuracy: 84%
Number of classes: 1
Dataset: 600k images
Accuracy: 95%
PeopleNet
TrafficCamNet
VehicleTypeNet
DashCamNet FaceDetect-IR
VehicleMakeNet
Highly Accurate | Re-Trainable | Out of Box Deployment
50. 50
BRIDGING THE REALITY GAP
https://arxiv.org/pdf/1804.06516.pdf NVR+University of Toronto
“leveraging the power of synthetic data improves upon results obtained using real data alone”
52. 52
Oxford Nanopore
Sequence Viral
Genome in 7Hrs
Plotly, NVIDIA
Real-Time
Infection Rate Analysis
ORNL, Scripps
Screen
2B Drug Compounds in
1 Day vs 1 Year
Structura, NIH, UT Austin
CryoSPARC
1st 3D Structure of Virus Spike Protein
NIH, NVIDIA
AI COVID-19
Classification
Kiwibot
Robot Medical Supply
Delivery
Whiteboard Coordinator
AI Elevated Body Temp
Screening System
NVIDIA SCIENTIFIC COMPUTING FIGHTS COVID-19
Data
Analytics
Simulation &
Visualization
AI Edge
53. 53
REVOLUTIONIZING
GENOMIC TESTING
Genome sequencing is the process of identifying
the DNA sequence of anything biological―animal,
food, water, air, plants. This ‘decoding’ is an
important step toward better scientific
understanding. Oxford Nanopore Technologies
is democratizing genome sequencing with
MinIT, a portable rapid analysis and device
control accessory to run the MinION
DNA/RNA sequencer. Powered by the
NVIDIA AGX System for high-throughput,
real-time analysis, MinIT will enable
DNA/RNA sequencing by anyone,
anywhere.
55. 55
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
nvidia.com/DLI
56. 56
NVIDIA
INCEPTION
PROGRAM
Accelerates AI startups with a boost of
GPU tools, tech and deep learning expertise
Startup Qualifications
Driving advances in the field of AI
Business plan
Incorporated
Web presence
Technology
DL startup kit*
Pascal Titan X
Deep Learning Institute (DLI) credit
Connect with a DL tech expert
DGX-1 ISV discount*
Software release notification
Live webinar and office hours
*By application
Marketing
Inclusion in NVIDIA marketing efforts
GPU Technology Conference (GTC)
discount
Emerging Company Summit (ECS)
participation+
Marketing kit
One-page story template
eBook template
Inception web badge and banners
Social promotion request form
Event opportunities list
Promotion at industry events
GPU ventures+
+By invitation
www.nvidia.com/inception
57. INNOVATION NEVER SLEEPS
October 5 - 9, 2020 Register at www.nvidia.com/GTC
EXPERIENCE WHAT’S NEXT AT NVIDIA GTC ONLINE
This autumn’s GTC will run continuously for five days, across seven
time zones, showcasing the latest breakthroughs in AI, HPC,
graphics, networking, and more.
Attend live events in the time zone that works best for you or browse
an extensive catalog of on-demand content showcasing innovative
uses of GPU technology.
Use code: NVALOWNDES