PRESENTATION

Q3 FY2024 — Presentation

Q3 FY2024PDF · 65 pages9.6 MB
Investor Presentation
Q3 FY24
November 27, 2023

-- 1 of 65 --

Except for the historical information contained herein, certain matters in this presentation including, but not limited to, statements as to: our financial position; our markets, market opportunity, demand and
growth drivers; a broadening set of GPU-specialized CSPs; entering the holidays with our best-ever line-up for gamers and creators; generative AI emerging as the new “killer app” for high-performance PCs;
being on track to exit the year at an annualized revenue run rate of $1 billion for our recurring software, support, and services offerings; AI emerging as a powerful demand driver for Professional Visualization;
Foxconn incorporating Omniverse into its manufacturing process; our financial outlook, and expected tax rates for the fourth quarter of fiscal 2024; our expectations of sequential growth to be driven by Data
Center, continued strong demand for compute and networking, and Gaming likely declining sequentially; the U.K. government building one of the world’s fastest AI supercomputers; Jülich building its next-gen
AI supercomputer; the combined AI compute capacity of all the supercomputers built on Grace Hopper across the U.S., EMEA and Japan next year; the benefits, impact, performance, features and availability of
our products and technologies; the benefits, impact, features and timing of our collaborations or partnerships; NVIDIA accelerated computing being broadly recognized as the way to advance computing as
Moore’s law ends and AI lifts off; accelerated computing being needed to tackle the most impactful opportunities of our time; AI driving a platform shift from general purpose to accelerated computing, and
enabling new, never-before-possible applications; trillion dollars of installed global data center infrastructure transitioning to accelerated computing; broader enterprise adoption of AI and accelerated
computing under way; AI and accelerated computing making possible the next big waves of autonomous machines and industrial digitalization; a rapidly growing universe of applications and industry innovation;
AI’s ability to augment creativity and productivity; generative AI as the most important computing platform of our generation; data centers becoming AI factories; full-stack and data center scale acceleration
driving significant cost savings and workload scaling; the high ROI of high compute performance; our belief that every important company will run its own AI factories; our dividend program plan; AI factories
expanding our market opportunity; our Automotive design win pipeline, ramp and production expectations; our aim to engage manufacturing suppliers and goal of effecting supplier adoption of science-based
environmental targets by fiscal 2026; and our plan for 100% renewable electricity for our operations and data centers by fiscal 2025 and annually thereafter are forward-looking statements.
These forward-looking statements and any other forward-looking statements that go beyond historical facts that are made in this presentation are subject to risks and uncertainties that may cause actual
results to differ materially. Important factors that could cause actual results to differ materially include: global economic conditions; our reliance on third parties to manufacture, assemble, package and test our
products; the impact of technological development and competition; development of new products and technologies or enhancements to our existing product and technologies; market acceptance of our
products or our partners' products; design, manufacturing or software defects; changes in consumer preferences and demands; changes in industry standards and interfaces; unexpected loss of performance
of our products or technologies when integrated into systems and other factors.
NVIDIA has based these forward-looking statements largely on its current expectations and projections about future events and trends that it believes may affect its financial condition, results of operations,
business strategy, short-term and long-term business operations and objectives, and financial needs. These forward-looking statements are subject to a number of risks and uncertainties, and you should not
rely upon the forward-looking statements as predictions of future events. The future events and trends discussed in this presentation may not occur and actual results could differ materially and adversely from
those anticipated or implied in the forward-looking statements. Although NVIDIA believes that the expectations reflected in the forward-looking statements are reasonable, the company cannot guarantee that
future results, levels of activity, performance, achievements or events and circumstances reflected in the forward-looking statements will occur. Except as required by law, NVIDIA disclaims any obligation to
update these forward-looking statements to reflect future events or circumstances. For a complete discussion of factors that could materially affect our financial results and operations, please refer to the
reports we file from time to time with the SEC, including our most recent Annual Report on Form 10-K, Quarterly Reports on Form 10-Q, and Current Reports on Form 8-K. Copies of reports we file with the SEC
are posted on our website and are available from NVIDIA without charge.
Many of the products and features described herein remain in various stages and will be offered on a when-and-if-available basis. The statements within are not intended to be, and should not be interpreted as
a commitment, promise, or legal obligation, and the development, release, and timing of any features or functionalities described for our products is subject to change and remains at the sole discretion of
NVIDIA. NVIDIA will have no liability for failure to deliver or delay in the delivery of any of the products, features or functions set forth herein.
NVIDIA uses certain non-GAAP measures in this presentation including non-GAAP gross profit, non-GAAP gross margin, non-GAAP operating expenses, non-GAAP operating income, non-GAAP operating margin,
non-GAAP net income, non-GAAP diluted earnings per share, and free cash flow. NVIDIA believes the presentation of its non-GAAP financial measures enhances investors' overall understanding of the company's
historical financial performance. The presentation of the company's non-GAAP financial measures is not meant to be considered in isolation or as a substitute for the company's financial results prepared in
accordance with GAAP, and the company's non-GAAP measures may be different from non-GAAP measures used by other companies. Further information relevant to the interpretation of non-GAAP financial
measures, and reconciliations of these non-GAAP financial measures to the most comparable GAAP measures, may be found in the slide titled “Reconciliation of Non-GAAP to GAAP Financial Measures”.

-- 2 of 65 --

• Q3 FY24 Earnings Summary
• Key Announcements This Quarter
• NVIDIA Overview
• Financials
• Reconciliation of Non-GAAP to GAAP Financial Measures
Content

-- 3 of 65 --

Q3 FY24
Earnings Summary

-- 4 of 65 --

Highlights
Record quarter driven by strong Data Center growth
• Total revenue up 206% Y/Y to $18.12B, well above outlook of $16.00B +/- 2%
• Data Center up 279% Y/Y to $14.51B
• Gaming up 81% Y/Y to $2.86B
Record Data Center revenue driven by continued ramp of NVIDIA HGX platform and InfiniBand networking
• Consumer internet and enterprise companies drove exceptional sequential growth, outpacing total growth
• Strong demand from all hyperscale cloud service providers (CSPs), and a broadening set of GPU-specialized CSPs
• Inference is contributing significantly to NVIDIA Data Center demand as AI is now in full production
Gaming growth reflects strong demand for GeForce RTX 40 series GPUs for back-to-school and the holidays
• GeForce RTX available at price points as low as $299 — entering the holidays with best-ever line-up for gamers and creators
• Gaming has doubled relative to pre-COVID levels even against the backdrop of lackluster PC market performance
• Gen AI emerging as new “killer app” for high-performance PCs — NVIDIA RTX is the natural platform for AI-application developers

-- 5 of 65 --

Q3 FY24 Financial Summary
All dollar figures are in millions other than EPS. Refer to Appendix for reconciliation of Non-GAAP measures.
GAAP Non-GAAP
Q3 FY24 Y/Y Q/Q Q3 FY24 Y/Y Q/Q
Revenue $18,120 +206% +34% $18,120 +206% +34%
Gross Margin 74.0% +20.4 pts +3.9 pts 75.0% +18.9 pts +3.8 pts
Operating
Income $10,417 +1,633% +53% $11,557 +652% +49%
Net Income $9,243 +1,259% +49% $10,020 +588% +49%
Diluted EPS $3.71 +1,274% +50% $4.02 +593% +49%
Cash Flow
from Ops $7,333 +1,771% +16% $7,333 +1,771% +16%
$5,931 $6,051
$7,192
$13,507
$18,120
56.1%
66.1%
66.8%
71.2%
75.0%
50.0%
55.0%
60.0%
65.0%
70.0%
75.0%
80.0%
85.0%
90.0%
1,500
3,500
5,500
7,500
9,500
11,500
13,500
15,500
17,500
19,500
Q3 FY23 Q4 FY23 Q1 FY24 Q2 FY24 Q3 FY24
Revenue($M) Non-GAAP GM

-- 6 of 65 --

Highlights
• Data Center compute revenue quadrupled from last year,
Networking revenue nearly tripled
• Strong, broad-based demand for NVIDIA accelerated computing
fueled by investment in the buildout of infrastructure for LLMs,
recommendation engines, and gen AI applications
• Networking business now exceeds a $10 billion annualized
revenue run rate
• NVIDIA H100 Tensor Core GPU instances are now generally
available in virtually every cloud, and are in high demand
• Vast majority of revenue driven by NVIDIA Hopper HGX, with a
lower contribution from the prior-gen Ampere GPU architecture
• New L40S GPU began to ship; first revenue quarter for GH200
• On track to exit the year at an annualized revenue run rate of $1
billion for our recurring software, support, and services offerings
Data Center
Revenue ($M)
$3,833 $3,616
$4,284
$10,323
$14,514
Q3 FY23 Q4 FY23 Q1 FY24 Q2 FY24 Q3 FY24
279% Y/Y
and
41% Q/Q

-- 7 of 65 --

Highlights
• Strong demand in the important back-to-school shopping season
• The RTX ecosystem continues to grow; there are now over 475
RTX enabled games and applications
• Released TensorRT-LLM for Windows, which speeds on-device
LLM inference by up to 4X
• GeForce NOW surpassed 1,700 PC titles including Alan Wake II,
Baldur’s Gate 3, Cyberpunk 2077: Phantom Liberty, and Starfield
Gaming
Revenue ($M)
$1,574
$1,831
$2,240
$2,486
$2,856
Q3 FY23 Q4 FY23 Q1 FY24 Q2 FY24 Q3 FY24
81% Y/Y
and
15% Q/Q

-- 8 of 65 --

$200
$226
$295
$379
$416
Q3 FY23 Q4 FY23 Q1 FY24 Q2 FY24 Q3 FY24
Highlights
• AI emerging as a powerful demand driver, including inference
for AI imaging in healthcare, edge AI in smart spaces and the
public sector
• Launched a new line of desktop workstations based on
NVIDIA RTX Ada Lovelace generation GPUs and ConnectX
SmartNICs
• Mercedes-Benz is using Omniverse-powered digital twins to
plan, design, build and operate its manufacturing and
assembly facilities
• Foxconn will incorporate Omniverse into its manufacturing
process
• Announced two new Omniverse Cloud services on Microsoft
Azure — for virtual factory simulation and autonomous
vehicle simulation
Professional Visualization
Revenue ($M)
108% Y/Y
and
10% Q/Q

-- 9 of 65 --

$251
$294 $296
$253 $261
Q3 FY23 Q4 FY23 Q1 FY24 Q2 FY24 Q3 FY24
Highlights
• Growth primarily driven by continued growth in self-driving
platforms based on NVIDIA DRIVE Orin SoC, and the ramp of
AI cockpit solutions with global OEM customers
• Extended automotive partnership with Foxconn to include
NVIDIA DRIVE Thor, next-generation automotive SoC
Automotive
Revenue ($M)
4% Y/Y
and
3% Q/Q

-- 10 of 65 --

$392
$2,249
$2,911
$6,348
$7,333
Q3 FY23 Q4 FY23 Q1 FY24 Q2 FY24 Q3 FY24
Highlights
• Y/Y and Q/Q growth primarily driven by higher revenue partially
offset by higher cash tax payments
• Utilized cash of $3.9 billion towards shareholder returns, including
$3.8 billion in share repurchases and $99 million in cash dividends
• Invested $291M in capex (includes principal payments on PP&E)
• Ended the quarter with $18.3B in gross cash and $9.8B in debt;
$8.5B in net cash
Sources & Uses of Cash
Cash Flow from Operations ($M)
Gross cash is defined as cash/cash equivalents & marketable securities.
Debt is defined as principal value of debt.
Net cash is defined as gross cash less debt.
1,771% Y/Y
and
16% Q/Q

-- 11 of 65 --

Q4 FY24 Outlook
Revenue $20.0 billion, plus or minus 2%
Expect strong Q/Q growth to be driven by Data Center, with continued strong demand for both compute
and networking. Gaming will likely decline Q/Q, as it is now more aligned with notebook seasonality
Gross Margins 74.5% GAAP and 75.5% non-GAAP, plus or minus 50 basis points
Operating Expense Approximately $3.17 billion GAAP and $2.20 billion non-GAAP
Other Income & Expense Income of approximately $200 million for GAAP and non-GAAP
Excluding gains and losses on non-affiliated investments
Tax Rate 15.0% GAAP and non-GAAP, plus or minus 1%, excluding discrete items
Refer to Appendix for reconciliation of Non-GAAP measures.

-- 12 of 65 --

Key Announcements
This Quarter

-- 13 of 65 --

New TensorRT-LLM Software
More Than Doubles Inference
Performance
• NVIDIA developed TensorRT-LLM, an open-source software
library that enables customers to more than double the
inference performance of their GPUs
• TensorRT-LLM on H100 GPUs provides up to an 8X
performance speedup compared to prior generation A100
GPUs running GPT-J 6B without the software
• 5.3X reduction in TCO and 5.6X reduction in energy costs
• With TensorRT-LLM for Windows, LLMs and generative AI
applications can run up to 4x faster locally on PCs and
Workstations powered by NVIDIA GeForce RTX and NVIDIA
RTX GPUs
• TensorRT-LLM for data centers now publicly available;
TensorRT-LLM for Windows in beta
1x
4x
8x
0X
1X
2X
3X
4X
5X
6X
7X
8X
A100 H100 August H100 TensorRT-
LLM
8X Increase in GPT-J 6B Inference Performance
1X
2.6X
4.6X
0X
1X
2X
3X
4X
5X
A100 H100 August H100 TensorRT-
LLM
4.6X Higher Llama2 Inference Performance
TensorRT-LLM Supercharges Hopper Performance
Software optimizations double leading performance
Text summarization, variable input/output length, CNN / DailyMail dataset | A100 FP 16 PyTorch eager mode / H100 FP8 | H100 FP8, TensorRT-LLM, in-flight batching

-- 14 of 65 --

NVIDIA Partners With Foxconn
to Build Factories and Systems
for the AI Industrial Revolution
• Foxconn, the world’s largest manufacturer, will integrate
NVIDIA technology to develop “AI factories”, a new class
of data centers
• Based on the NVIDIA accelerated computing platform,
including NVIDIA GH200 and NVIDIA AI Enterprise
software, these AI factories will power a wide range of
applications, including:
• Digitalization of manufacturing and inspection workflows
• Development of AI-powered EVs and robotics platforms
• A growing number of language-based generative AI services
• In addition:
• Foxconn Smart EV will be built on NVIDIA DRIVE Hyperion 9,
next-gen platform for autonomous automotive fleets,
powered by NVIDIA DRIVE Thor, our future automotive SoC
• Foxconn Smart Manufacturing robotic systems will be built
on the NVIDIA Isaac autonomous mobile robot platform.
• Foxconn Smart City will incorporate the NVIDIA Metropolis
intelligent video analytics platform AI factories are a new class of data centers, optimized for
refining data and training, inferencing, and generating AI

-- 15 of 65 --

NVIDIA Partners With India Tech
Giants to Advance AI Across
World’s Most Populous Nation
NVIDIA announced collaborations with Reliance Industries,
Tata Group and Infosys to bring AI technology and skills to India
• With Reliance, the companies will work together to develop
India’s own foundation LLM trained on India’s diverse
languages and tailored for generative AI applications; build
supercomputing infrastructure to support the exponential
computational demands of AI
• With Tata, the collaboration will bring a state-of-the-art
AI supercomputer to provide infrastructure-as-a-service
and platform for AI services in India
• With Infosys, the partnership will bring the NVIDIA AI Enterprise
ecosystem of models, tools, runtimes and GPU systems to drive
productivity gains with generative AI applications and solutions
• Infosys plans to set up an NVIDIA Center of Excellence where it will
train and certify 50,000 of its employees on NVIDIA AI technology

-- 16 of 65 --

NVIDIA Sets New LLM Training
Record With Largest MLPerf
Submission Ever
• NVIDIA set six new performance records in this round,
with the performance increase stemming from a combination
of advances in software and scaled-up hardware
• 2.8x faster on generative AI – completing a training benchmark
based on a GPT-3 model with 175 billion parameters trained on
1 billion tokens in just 3.9 minutes
• 1.6x faster on training recommender models
• 1.8x faster on training computer vision models
• The GPT-3 benchmark ran on NVIDIA Eos – a new AI
supercomputer powered by 10,752 H100 GPUs and NVIDIA
Quantum-2 InfiniBand networking
• The 10,752 H100 GPUs far surpassed the scaling in AI training
in June, when NVIDIA used 3,584 Hopper GPUs
• The 3x scaling in GPU numbers delivered a 2.8x scaling in
performance, a 93% efficiency rate thanks in part to software
optimizations
• Microsoft Azure achieved similar results on a nearly identical
cluster, demonstrating the efficiency of NVIDIA AI in public
cloud deployments
Stable Diffusion
2.5 Minutes
New Workload
GPT-3 175B (1B Tokens)
3.9 Minutes
2.8X Faster
DLRM-dcnv2
1 Minute
1.6X Faster
BERT-Large
7.2 Seconds
1.1X Faster
RetinaNet
55.2 Seconds
1.8X Faster
3D U-Net
46 Seconds
1.07X Faster
MLPerf™ Training v3.1. Results retrieved from www.mlperf.org on November 8, 2023. Format: Chip Count, MLPerf ID | GPT-3:
3584x 3.0-2003, 10752x 3.1-2007 | Stable Diffusion: 1024x 3.1-2050 | DLRMv2: 128x 3.0-2065, 128x 3.1-2051 | BERT-Large:
3072x 3.0-2001, 3472x 3.1-2053 |
RetinaNet: 768x 3.0-2077, 2048x 3.1-2052 | 3D U-Net: 432x 3.0-2067, 768x 3.1-2064. The MLPerf™ name and logo are
trademarks of MLCommons Association in the United States and other countries. All rights reserved. Unauthorized use
strictly prohibited. See www.mlcommons.org for more information.
Six New Performance Records
The fastest gets even faster

-- 17 of 65 --

New NVIDIA HGX H200
Supercharges Hopper
• NVIDIA H200 is the first GPU to offer HBM3e — faster,
larger memory to fuel the acceleration of generative AI
and large language models, while advancing scientific
computing for HPC workloads
• H200 delivers 141GB of memory at 4.8 terabytes per
second, nearly double the capacity and 2.4X more
bandwidth compared with its predecessor, NVIDIA A100
• Boosts inference speed by up to 2X compared to H100
GPUs when handling LLMs such as Llama2
• Microsoft announced plans to add the H200 to Azure next
year for larger model inference with no increase in latency
• H200-powered systems from the world’s leading server
manufacturers and cloud service providers are expected
to begin shipping in the second quarter of 2024

-- 18 of 65 --

Grace Hopper Gains Significant
Traction with Supercomputing
Customers
• Initial shipments to Los Alamos National Lab and the Swiss
National Supercomputing Centre took place in the third
quarter
• The U.K. government announced it will build one of the
world’s fastest AI supercomputers with almost 5.5K Grace
Hopper Superchips
• German supercomputing center Jülich will build its next-
gen AI supercomputer, with close to 24K Grace Hopper
Superchips and Quantum-2 InfiniBand
• Will be the world’s most powerful AI system with over 90
exaflops of AI performance
• Marks the debut of a quad NVIDIA GH200 Grace Hopper
Superchip node configuration
• Combined AI compute capacity of all the supercomputers
built on Grace Hopper across the U.S., EMEA and Japan
next year estimated to exceed 200 exaflops
Cumulative AI FLOPS
0
50
100
150
200
250
300
350
400
2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025
Delta
Cumulative AI Performance
(ExaFLOPS of AI)
Polaris
Perlmutter Leonardo
Pangea Selene
Tsubame3Piz Daint Summit
MareNostrum-5
Eos
Azure ND H100 v5
Alps
Vista
Venado
Jupiter
OFP-II
JUPITER
Isambard AI

-- 19 of 65 --

NVIDIA AI Foundry Service for
Enterprises on Microsoft Azure
• Introduced new NVIDIA AI foundry service for the
development and tuning of custom generative AI
enterprise applications, running on Microsoft Azure
• Customers can bring their domain knowledge and
proprietary data, and we help them build their AI models
using our AI expertise and software stack in DGX Cloud
AI factory – all with enterprise-grade security and support
• Businesses can deploy their customized models with the
NVIDIA AI Enterprise software runtime to power generative
AI applications such as intelligent search, summarization,
and content generation
• Industry leaders SAP SE, Amdocs and Getty Images are
among the first customers of NVIDIA AI foundry service
Create from
Foundation Model
AI Foundations NeMo DGX Cloud
Run
Anywhere
Vector Store
LLM
Agent LLM
Prompts
RAG
Your Enterprise
Model
Running on
NVIDIA AI Enterprise

-- 20 of 65 --

NVIDIA Spectrum-X Ethernet
networking platform for AI Available
Soon from Dell, HPE and Lenovo
• Purpose-built for gen AI, Spectrum-X offers enterprises
a new class of Ethernet networking that can achieve 1.6x
higher networking performance for AI communication
versus traditional Ethernet offerings
• Dell, Hewlett Packard Enterprise and Lenovo will be the
first to integrate NVIDIA Spectrum-X Ethernet networking
technologies for AI into their server lineups
• New systems bring together Spectrum-X with NVIDIA
GPUs, NVIDIA AI Enterprise software and NVIDIA AI
Workbench software to provide enterprises the building
blocks to transform their businesses with generative AI
• Available in the first quarter of next year

-- 21 of 65 --

NVIDIA Collaborates With
Genentech to Accelerate Drug
Discovery Using Generative AI
• Genentech is pioneering the use of generative AI to
discover and develop new therapeutics and deliver
treatments to patients more efficiently
• NVIDIA will work with Genentech to accelerate
Genentech’s proprietary algorithms on NVIDIA DGX Cloud
• Genentech plans to use NVIDIA BioNeMo to help
accelerate and optimize their AI drug discovery platform
• NVIDIA plans to use insights learned from this
collaboration to improve its BioNeMo platform
• BioNeMo is now generally available as a training service

-- 22 of 65 --

NVIDIA Overview

-- 23 of 65 --

24
Headquarters: Santa Clara, CA
NVIDIA pioneered accelerated computing to help solve impactful
challenges classical computers cannot. A quarter of a century in the
making, NVIDIA accelerated computing is broadly recognized as the
way to advance computing as Moore’s law ends and AI lifts off.
NVIDIA’s platform is installed in several hundred million computers,
is available in every cloud and from every server maker, powers 76%
of the TOP500 supercomputers, and boasts 4.5 million developers.

-- 24 of 65 --

NVIDIA’s Accelerated Computing Platform
Full-stack innovation across silicon, systems and software
cuNumeric CV-CUDA cuQuantum Parabricks Sionna Jetpack
RAPIDS Spark cuDNN cuGraph TensorRT Triton Deepstream Flare
DOCA Mag IO Aerial
NVIDIA HPC NVIDIA AI NVIDIA Omniverse
RTX DGX HGX EGX OVX AGX
Super
POD IGX
DPU	CPU	GPU
3-CHIPS
PLATFORMS
AI APPLICATION FRAMEWORK
ACCELERATION
LIBRARIES
CLOUD-TO-EDGE
DATACENTER-TO-
ROBOTIC SYSTEMS
With nearly three decades of singular focus,
NVIDIA is expert at accelerating software
and scaling compute by a
Million-X, going well beyond Moore’s law
Accelerated computing requires full-stack
innovation — optimizing across every layer
of computing — from silicon and systems to
software and algorithms, demanding deep
understanding of the problem domain
Our full-stack platforms — NVIDIA HPC,
NVIDIA AI, and NVIDIA Omniverse —
accelerate high performance computing,
AI and industrial digitalization workloads
We accelerate workloads at data center
scale, across thousands of compute nodes,
treating the network and storage as part of
the computing fabric
Our platform extends from the cloud and
enterprise data centers to supercomputing
centers, edge computing and PCs

-- 25 of 65 --

What Is Accelerated Computing?
Not just a superfast chip – accelerated computing
is a full-stack combination of:
• Chip(s) with specialized processors
• Algorithms in acceleration libraries
• Domain experts to refactor applications
To speed-up compute-intensive parts of an application
A full-stack approach: silicon, systems, software
For example:
• If 90% of the runtime can be accelerated by 100X,
the application is sped up 9X
• If 99% of the runtime can be accelerated by 100X,
the application is sped up 50X
• If 80% of the runtime can be accelerated by 500X,
or even 1000X, the application is sped up 5X
Amdahl’s law:
The overall system speed-up (S) gained by optimizing a
single part of a system by a factor (s) is limited by the
proportion of execution time of that part (p).
𝑆 = 1
1 − 𝑝 + 𝑝
𝑠

-- 26 of 65 --

Why Accelerated Computing?
Accelerated computing is needed to tackle the most
impactful opportunities of our time—like AI, climate
simulation, drug discovery, ray tracing, and robotics
NVIDIA is uniquely dedicated to accelerated computing
—working top-to-bottom, refactoring applications and
creating new algorithms, and bottom-to-top—inventing
new specialized processors, like RT Core and Tensor Core
“It’s the end of Moore’s Law as we know it.”
- John Hennessy Oct 23, 2018
“Moore’s Law is dead.”
- Jensen Huang, GTC 2013
Advancing computing in the post-Moore’s Law era
102
103
104
105
106
107
109
108
1980 1990 2000 2010 2020 2030
GPU-Computing perf
2X per year 1000X
In 10 years
Single-threaded CPU perf
1.5X perf per year
1.1X per year
Trillions of Operations per Second (TOPS)

-- 27 of 65 --

A new computing era has begun
Accelerated computing enabled the rise of AI,
which is driving a platform shift from general
purpose to accelerated computing, and enabling
new, never-before-possible applications
The trillion dollars of installed global data center
infrastructure will transition to accelerated
computing to achieve better performance, energy-
efficiency and cost by an order of magnitude
Hyperscale cloud service providers and consumer
internet companies have been the early adopters
of AI and accelerated computing, with broader
enterprise adoption now under way
AI and accelerated computing will also make possible
the next big waves — autonomous machines and
industrial digitalization
Waves of Adoption of Accelerated Computing
A generational computing platform shift
Cloud Service Providers
& Consumer Internet
Enterprise
Autonomous Vehicles
& Robotics
Industrial Digitalization

-- 28 of 65 --

NVIDIA Accelerated Computing for Every Wave
NVIDIA HGX is an AI supercomputing platform purpose-built for AI. It includes 8 NVIDIA GPUs, as well as
interconnect and networking technologies, delivering order-of-magnitude performance speed-ups for AI
over CPU servers. It is broadly available from all major server OEMs/ODMs. NVIDIA DGX, an AI server based
on the same architecture, along with NVIDIA AI software and support, is also available
NVIDIA Omniverse is a software platform for designing,
building, and operating 3D and virtual world simulations.
It harnesses the power of NVIDIA graphics and AI technologies
and runs on NVIDIA-powered data centers and workstations
NVIDIA DRIVE is a full-stack platform for autonomous vehicles (AV) that
includes hardware for in-car compute, such as the Orin system-on-chip, and
the full AV and AI cockpit software stack
NVIDIA AI Enterprise is the operating system of AI, with enterprise-grade security, stability,
manageability and support. It is available on all major CSPs and server OEMs and supports
enterprise deployment of AI in production
NVIDIA DGX Cloud is a cloud service that allows enterprises immediate access to the
infrastructure and software needed to train advanced models for generative
AI and other groundbreaking applications
Cloud Service Providers
& Consumer Internet
Enterprise
Autonomous Vehicles
& Robotics
Industrial Digitalization

-- 29 of 65 --

The NVIDIA accelerated computing platform has attracted
the largest ecosystem of developers, supporting a rapidly
growing universe of applications and industry innovation
Developers can engage with NVIDIA through CUDA —
our parallel computing programming model introduced
in 2006 — or at higher layers of the stack, including libraries,
pre-trained AI models, SDKs and other development tools
NVIDIA’s Accelerated Computing Ecosystem
Developers CUDA Downloads*
AI Startups GPU-Accelerated
Applications
2020 2023
6K
15K
2020 2023
4.5M
1.8M
2020 2023
48M
20M
2020 2023
3,200
700
300 Libraries
600 AI Models
100 Updated in the Last Year
*Cumulative

-- 30 of 65 --

The virtuous cycle of NVIDIA’s accelerated computing starts with an
installed base of several hundred million GPUs, all compatible with the
CUDA programming model
• For developers — NVIDIA’s one architecture and large installed base
give developer’s software the best performance and greatest reach
• For end users — NVIDIA is offered by virtually every computing
provider and accelerates the most impactful applications from
cloud to edge
• For cloud providers and OEMs — NVIDIA’s rich suite of Acceleration
Platforms lets partners build one offering to address large markets
including media & entertainment, healthcare, transportation, energy,
financial services, manufacturing, retail, and more
• For NVIDIA — Deep engagement with developers, computing
providers, and customers in diverse industries enables unmatched
expertise, scale, and speed of innovation across the entire
accelerated computing stack — propelling the flywheel
NVIDIA’s Multi-Sided Platform and Flywheel
NVIDIA Accelerated Computing Virtuous Cycle
Cloud & OEMs	Developers
Installed Base
Scale
End-Users
Speed-Up
R&D $

-- 31 of 65 --

Search & Social Media	Office AI Copilots
$700B in digital advertising annually	Over 1B knowledge workers
AI Content Creation
50M creators globally
Legal Services, Education
1M legal professionals in the US
9M educators in the US
Financial Services	AI Software Development
678B annual credit card transactions	30M software developers globally
Customer Service with AI
15M call center agents globally
Drug Discovery
1018 molecules in chemical space
40 exabytes of genome data
Agri-Food | Climate
1B people in agri-food worldwide
Earth-2 for km-scale simulation
Knowledge workers will use copilots based on large
language models to generate documents, answer
questions, or summarize missed meetings, emails
and chats — adding hours of productivity per week
Copilots specialized for fields such as software
development, legal services or education can boost
productivity by as much as 50%
Social media, search and e-commerce apps are
using deep recommenders to offer more relevant
content and ads to their customers, increasing
engagement and monetization
Creators can generate stunning, photorealistic
images with a single text prompt — compressing
workflows that take days or weeks into minutes in
industries from advertising to game development
Call center agents augmented with AI chatbots
can dramatically increase productivity and
customer satisfaction
Drug discovery, financial services, agriculture and
food services and climate forecasting are seeing
order-of-magnitude workflow acceleration from AI
Huge ROI from AI Driving a Powerful New Investment Cycle
AI can augment creativity and productivity by orders of magnitude across industries
Source: Goldman Sachs, Cowen, Statista, Capital One, Wall Street Journal, Resource Watch, NVIDIA internal analysis

-- 32 of 65 --

The era of generative AI has arrived, unlocking new
opportunities for AI across many different applications
Generative AI is trained on large amounts of data
to find patterns and relationships, learning the
representation of almost anything with structure
It can then be prompted to generate text, images,
video, code, or even proteins
For the very first time, computers can augment the
human ability to generate information and create
1,600+ Generative AI companies are building on NVIDIA
Generative AI
The most important computing platform of our generation
TEXT
AUDIO
IMAGE
3D
VIDEO
DNA
PROTEIN
MOLECULE
ANIMATION
ANIMATION
MOLECULE
PROTEIN
DNA
VIDEO
3D
IMAGE
AUDIO
TEXT

-- 33 of 65 --

Large Language Models, based on the Transformer architecture,
are one of today’s most important advanced AI technologies,
involving up to trillions of parameters that learn from text.
Developing them is an expensive, time-consuming process that
demands deep technical expertise, distributed data center-scale
infrastructure, and a full-stack accelerated computing approach.
Modern AI is a Data Center Scale Computing Workload
Data centers are becoming AI factories: Data as input, intelligence as output
AI Training Computational Requirements Fueling Giant-Scale AI Infrastructure
NVIDIA compute & networking GPU | DPU | CPU
AlexNet
VGG-19
Seq2Seq Resnet
InceptionV3
Xception
ResNeXt
DenseNet201 ELMo
MoCo ResNet50
Wav2Vec 2.0
Transformer
GPT-1
BERT Large
GPT-2
XLNet
Megatron-NLG
Microsoft T-NLG
GPT-3
MT NLG 530B
BLOOM
Chinchilla
PaLM
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023
Training Compute (petaFLOPs)
Before Transformers = 8X / 2yrs
Transformers = 215X / 2yrs
102
103
104
105
106
107
108
109
1010

-- 34 of 65 --

Full-Stack & Data Center Scale Acceleration
Drive significant cost savings and workload scaling
Classical Computing—960 CPU-only servers Accelerated Computing—2 GPU servers
25X lower cost
84X better energy-efficiency
Application Application
Re-Engineered for Acceleration
Magnum IO
CUDA-X Acceleration Libraries	CPU server racks
LLM Workload: Bert-Large Training and Inference | CPU Server: Dual-EYPC 7763 | GPU Server: Dual-EPYC 7763 + 8X H100 PCIe GPUs

-- 35 of 65 --

The High ROI of High Compute Performance
Rental
Cost
4-Year Cost of AI Infrastructure
~$1B
16K GPU
4-Year Rental Opportunity
@$4 per GPU-HR
~$2.5B
GPU Compute
Networking
DC Facility Build
& Operate
25%
Performance
Increase Worth
$600M+
15% Utilization
Increase Worth
$350M+
$1 upfront investment in NVIDIA compute and networking can translate to $5 in CSP revenue over 4 years
Illustrative example of NVIDIA GPU cost vs AI infrastructure total cost of ownership (TCO)

-- 36 of 65 --

Training & Inference — One Architecture
Cloud | On-Prem | Edge
Training Inference
IN THE DATA CENTER
NVIDIA L40
Image Generation
NVIDIA L4
AI Video
NVIDIA H100 | L40S
Universal GPUs
NVIDIA Grace Hopper
RecSys, Gen AI
AGX
Functionally-Safe System
for Autonomous Vehicles
IGX
Industrial-Grade System
for Healthcare, Logistics,
Manufacturing
AT THE EDGE	NVIDIA DGX | HGX H100
NVIDIA L40S

-- 37 of 65 --

Powering the AI Industrial Revolution
Building and Running Enterprise Gen AI Applications
Enterprise
AI Chatbot
with “RAG”
NVIDIA AI foundry service
for building Enterprise AI applications
NVIDIA AI enterprise ecosystem
for running Enterprise AI applications
Enterprise AI chatbots
Are built with Retrieval
Augmented Generation (RAG),
which augments the knowledge
in the LLM with Enterprise data
mapped to a Vector Database,
thus reducing “hallucinations”.
Developers can connect
additional or 3rd party services to
the AI chatbot via cloud AI APIs.
Vector Database
Custom LLM
Model Container
NVIDIA AI
Enterprise
Cloud AI
APIs
AI
Foundation
Model
Tech
DGX Cloud
Factory
NVPS
Experts
NVIDIA
DGX Cloud
NVIDIA AI Foundation
Pre-Trained LLMs
Enterprise SaaS
& AI Platforms
Enterprise On-Prem
Cloud
DGX Cloud

-- 38 of 65 --

The NVIDIA AI Foundry
Model on DGX Cloud
NVIDIA’s “AI foundry” service leverages our AI infrastructure and
expertise to build custom AI models for enterprise customers —
analogous to a semiconductor foundry that uses its infrastructure
and expertise to build custom chips for fabless customers.
An enterprise customer starts with an NVIDIA or 3rd party pre-trained
AI model, available in NVIDIA AI Foundations. This model making
service includes frameworks such as NVIDIA NeMo for custom LLMs
and NVIDIA Picasso for custom generative AI for visual design.
With help from NVIDIA experts, the enterprise customer fine-tunes
the model on their proprietary enterprise data and adds guardrails,
using tools available in NVIDIA AI Foundations.
The fine-tuning and optimization is done on NVIDIA DGX Cloud,
a cloud service that allows enterprises immediate access to NVIDIA
AI infrastructure and software, hosted at partner cloud providers.
The enterprise customer ends up with a fully-trained and optimized
AI model, fine-tuned on their proprietary enterprise data, that can be
deployed anywhere — in the cloud or on-prem.
The NVIDIA AI Foundry model generates revenue based on per-node,
per-month consumption of NVIDIA DGX Cloud.
For building enterprise AI applications
NVIDIA AI Foundations
NeMo | Picasso
Pre-trained LLMs
NVIDIA
DGX Cloud
NVIDIA AI foundry

-- 39 of 65 --

AI Factories — A New
Class of Data Centers
“AI factories” are a new class of data centers specially built for
processing, refining and transforming vast amounts of data into
valuable AI models and tokens.
Unlike traditional data centers built for IT workloads, AI factories
are built to deliver automated, professional skills.
AI factories are not multi-workload or multi-tenant. They run one
workload – an AI model – and have just one customer or owner —
analogous to a traditional factory.
AI factories can be built on-prem, in the cloud, or in the data centers
of SaaS and AI platform vendors.
We believe that in the future, every important company will run its
own AI factories in order to securely process its valuable proprietary
data and turn it into monetizable tokens, encapsulating its
knowledge, intelligence, and creativity.
In addition to the up-front revenue opportunity from data center
systems, NVIDIA can generate recurring revenue from AI factories
for their use of NVIDIA AI Enterprise, the operating system for
enterprise AI.
For running enterprise AI applications
DATA TOKENS
Enterprise SaaS
& AI Platforms
Enterprise
On-Prem
Cloud
AI Factory
NVIDIA AI Enterprise

-- 40 of 65 --

NVIDIA AI Enterprise
The operating system for enterprise AI
NVIDIA AI Enterprise is software for deploying and running AI with
enterprise-grade security, API stability, manageability and support.
Cloud-native and available in every major cloud marketplace.
Certified to run on servers and workstations from all major OEMs.
Supported by all major global system integrators.
Integrated with and distributed by VMware.
NVIDIA AI Enterprise
AI Use Cases and Workflows
LLM Speech AI Recommenders Cybersecurity
Medical
Imaging
Video
Analytics
Route
Optimization
More
… Consumption pricing
per GPU-hour
Subscription pricing
per GPU/year
(included with H100 PCIe/DGX)
NVIDIA AI Enterprise
NVIDIA Certified Server
Dell | HPE | Lenovo
Azure | GCP | OCI | AWS
Run Anywhere
Cloud

-- 41 of 65 --

GSI & Service Delivery
Software Platforms	AI Platforms
Public Cloud
Marketplaces
Server OEMs
NVIDIA AI Enterprise
Broad and deep ecosystem and distribution to reach every enterprise
Private Cloud

-- 42 of 65 --

NVIDIA Go-to-Market Across Cloud and On-Premises
Reaching customers everywhere
INFERENCE
Cloud On-Prem
MGX AGX IGX	HGX
DGX
NVIDIA AI Foundations - Cloud services for
customizing and operating generative AI models
DGX Cloud
Partners

-- 43 of 65 --

$11,716 $10,918
$16,675
$26,914
$38,819
$26,974
FY19 FY20 FY21 FY22 FY23 YTD FY24
45
41
7 3 4 19
75
321 Gaming
Data Center
ProViz
Auto
OEM & Other
Driving Strong & Profitable Growth
Revenue ($M)
Fiscal year ends in January. Refer to Appendix for reconciliation of Non-GAAP measures. Operating margins rounded to the nearest percent.
FY23 financial metrics reflect a $2.2B charge for inventory and related reserves primarily related to Data Center and Gaming.
YTD FY24	YTD FY21
$4,407 $3,735
$6,803
$12,690
$9,040
$22,385
38% 34%
41%
47%
34%
58%
30%
40%
50%
60%
70%
$0
$4,000
$8,000
$12,000
$16,000
$20,000
$24,000
FY19 FY20 FY21 FY22 FY23 YTD FY24
Operating Income (Non-GAAP, $M) Operating Margin (Non-GAAP)

-- 44 of 65 --

$7,233 $6,821
$10,947
$17,969
$15,965
$28,000
62% 63%
66%
67%
59%
72%
55%
60%
65%
70%
75%
80%
$0
$6,000
$12,000
$18,000
$24,000
$30,000
FY19 FY20 FY21 FY22 FY23 YTD FY24
Gross Profit (Non-GAAP, $M) Gross Margin (Non-GAAP)
Cost comparison example based on latest available NVIDIA A100 GPU and Intel CPU inference results in the commercially available category of
the MLPerf industry benchmark; includes related infrastructure costs such as networking.
NVIDIA Gross Margins Reflect Value of Acceleration
FY23 financial metrics reflect a $2.2B charge for inventory and related reserves primarily related to Data Center and Gaming.
Fiscal year ends in January. Refer to Appendix for reconciliation of Non-GAAP measures. Gross margins are rounded to the nearest percent.
Accelerated computing requires full-stack and data
center-scale innovation across silicon, systems,
algorithms and applications.
Significant expertise and effort are required, but
application speed-ups can be incredible, resulting
in dramatic cost and time-to-solution savings.
For example, 2 NVIDIA HGX nodes with 16 NVIDIA H100 GPUs
that cost $400K can replace 960 nodes of CPU servers that
cost $10M for the same LLM workload.
NVIDIA chips carry the value of the full-stack,
not just the chip.

-- 45 of 65 --

$3.1B
$4.3B $4.7B
$8.0B
$3.8B
$15.7B
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
FY19 FY20 FY21 FY22 FY23 YTD FY24
Strong Cash Flow Generation
Fiscal year ends in January. Refer to Appendix for reconciliation of Non-GAAP measures.
Share Repurchase
$10B repurchased in FY23
$25.2B Remaining Authorization as of end of Q3
Dividend
$398M in FY 2023
Plan to Maintain1
Strategic Investments
Growing Our Talent
Platform Reach & Ecosystem
Free Cash Flow (Non-GAAP) Capital Allocation
1 Subject to continuing determination by our Board of Directors.

-- 46 of 65 --

DGX/HGX/MGX/IGX systems
GPU | CPU | DPU | Networking
NVIDIA AI software
Our Market Platforms at a Glance
FY23 Revenue $15.0B
5-YR CAGR 51%
FY23 Revenue $9.1B
5-YR CAGR 10%
FY23 Revenue $1.5B
5-YR CAGR 11%
FY23 Revenue $0.9B
5-YR CAGR 10%
GeForce GPUs for PC gaming
GeForce NOW cloud gaming
DRIVE Hyperion sensor architecture
with AGX compute
DRIVE AV & IX full stack software
for ADAS, AV & AI cockpit
NVIDIA RTX GPUs
for workstations
Omniverse software
Professional Visualization Automotive
33% of FY23 Revenue 6% of FY23 Revenue 3% of FY23 Revenue
Data Center
56% of FY23 Revenue
Gaming

-- 47 of 65 --

$2,932 $2,983
$6,696
$10,613
$15,005
$29,121
FY19 FY20 FY21 FY22 FY23 YTD FY24
Data Center
The leading accelerated computing platform
51% 5-YR CAGR
Through FY23
Leader in AI & HPC
#1 in AI training and inference
Used by all hyperscale and major cloud computing
providers and 40,000 enterprises
Powers 76% of the TOP500 supercomputers
Growth Drivers
Broad data center platform transition from
general-purpose to accelerated computing
Emergence of “AI factory” — optimized for refining
data and training, inferencing, and generating AI
Broader and faster product launch cadence to meet
a growing and diverse set of AI opportunities
DGX Cloud services and NVIDIA AI Enterprise software
for building and running enterprise AI applications
Revenue ($M)

-- 48 of 65 --

NVIDIA AI — One Architecture | Train and Deploy Everywhere
One-Year Rhythm
2024	2023
CPU + GPU
GH200
GB200
H100
H200
Quantum
Spectrum-X
GPU
B100
X100
GX200
2025
L40S
B40
X40
GH200NVL
GB200NVL
GX200NVL
x86 Training & Inference
x86 Enterprise & Inference
InfiniBand AI Infrastructure
Ethernet Enterprise &
Hyperscale AI Infrastructure
400G
800G
1,600G
400G
800G
1,600G
Arm Training & Inference
Arm Inference
DPU
Enterprise & Hyperscale
Infrastructure Computing
BlueField-3
BlueField-4

-- 49 of 65 --

$6,246
$5,518
$7,759
$12,462
$9,067
$7,582
FY19 FY20 FY21 FY22 FY23 YTD FY24
Gaming
GeForce — the world’s largest gaming platform
10% 5-YR CAGR
Through FY23
Revenue ($M) Leader in PC Gaming
Strong #1 market position
15 of the top 15 most popular GPUs on Steam
Leading performance & innovation
200M+ gamers on GeForce
Growth Drivers
Rising adoption of NVIDIA RTX in games
Expanding universe of gamers & creators
Gaming laptops & Gen AI on PCs
GeForce NOW Cloud gaming

-- 50 of 65 --

GeForce Extends Growth, Large Upgrade Opportunity
More Gamers, Richer Mix Installed Base Needs Upgrade Ada: 3X Turing Ramp at $699+
FY20 FY23
3YR CAGR
ASP 10%
Units 9%
47% RTX
20% RTX3060+
Performance
$699+ Cumulative Sell-Through $
20% CAGR
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
NVIDIA
Ampere
NVIDIA
Ada
Weeks After Launch
NVIDIA
Turing
Source: NVIDIA estimates
GeForce Gaming Revenue Installed Base
RTX
3060+

-- 51 of 65 --

$1,130 $1,212
$1,053
$2,111
$1,544
$1,090
FY19 FY20 FY21 FY22 FY23 YTD FY24
Professional Visualization
Workstation graphics
Leader in Workstation Graphics
95%+ market share in graphics
for workstations
45M Designers and Creators
Strong software ecosystem with over 100 RTX
accelerated and supported applications
Growth Drivers
Ray Tracing and generative AI revolutionizing
design and content creation
Expanding universe of designers and creators
Collaborative 3D design / Omniverse
Hybrid work environments
11% 5-YR CAGR
Through FY23
Revenue ($M)

-- 52 of 65 --

$641 $700
$536 $566
$903 $810
FY19 FY20 FY21 FY22 FY23 YTD FY24
Automotive
Autonomous Vehicles (AV) and AI Cockpit
Leader in Autonomous Driving
NVIDIA DRIVE is our end-to-end Autonomous Vehicle (AV) and AI
Cockpit platform featuring a full software stack and is powered
by NVIDIA (systems-on-a-chip) SoCs in the vehicle
DRIVE Orin SoC ramp began in FY23
Next-generation DRIVE Thor SoC ramp to begin in FY26
Over 40 customers including 20 of top 30 EV makers,
7 of top 10 truck makers, 8 of top 10 robotaxi makers
Growth Drivers
Adoption of centralized car computing and
software-defined vehicle architectures
AV software and services:
Mercedes-Benz
Jaguar Land Rover
Revenue ($M)
10% 5-YR CAGR
Through FY23

-- 53 of 65 --

$1 Trillion Long-Term Available Market Opportunity
Cloud Service Providers &
Consumer Internet
Enterprise
Autonomous Vehicles
& Robotics
Industrial Digitalization
Data Center Systems
$300B
Omniverse Enterprise
$150B
Autonomous Machines
$300B
NVIDIA AI Enterprise & DGX Cloud
$150B
Gaming
$100B

-- 54 of 65 --

Financials

-- 55 of 65 --

Annual Cash & Cash Flow Metrics
Free Cash Flow (Non-GAAP) — $M Cash Balance — $M
Operating Income (Non-GAAP) — $M Operating Cash Flow — $M
3,743
4,761
5,822
9,108
5,641
FY19 FY20 FY21 FY22 FY23
4,407 3,735
6,803
12,690
9,040
FY19 FY20 FY21 FY22 FY23
3,143
4,272 4,677
8,049
3,750
FY19 FY20 FY21 FY22 FY23
7,422
10,897 11,561
21,208
13,296
FY19 FY20 FY21 FY22 FY23
Cash balance is defined as cash and cash equivalents plus marketable securities
Refer to Appendix for reconciliation of non-GAAP measures

-- 56 of 65 --

Corporate Responsibility
Time Magazine’s 100 Most Influential
Companies
Fast Company’s Best Workplaces for
Innovators
Fortune’s World’s Most Admired Companies
Wall Street Journal’s Management
Top 250 All-Stars
“America’s Most Just Companies”
CNBC
“100 Best Companies to Work For”
FORTUNE
“Most Responsible Companies”
NEWSWEEK
“Best Places to Work for LGBT Equality”
HUMAN RIGHTS CAMPAIGN
NVIDIA GPUs are typically 20X more energy efficient for
certain AI and HPC workloads than traditional CPUs
Plan to achieve & maintain 100% renewable electricity for our
operations and data centers by FY25 and annually thereafter
A Place For People To Do Their Life’s Work	Environmentally Conscious
By FY26, aim to engage manufacturing suppliers comprising at
least 67% of NVIDIA’s scope 3 category 1 GHG emissions with
goal of effecting supplier adoption of science-based targets
Management
43% of Board is Gender, Racially,
or Ethnically Diverse
93% of Directors are independent
Corporate Governance

-- 57 of 65 --

Reconciliation of Non-GAAP to
GAAP Financial Measures

-- 58 of 65 --

Reconciliation of Non-GAAP to GAAP Financial Measures
Non-GAAP
Acquisition-Related
and Other Costs
(A)
Stock-Based
Compensation
(B)
IP-Related
Costs
Other
(C)
Tax Impact of
Adjustments GAAP
Q3
FY24
Gross margin
($ in million)
$13,583 (119) (38) (26) — — $13,400
75.0% (0.7) (0.2) (0.1) — — 74.0%
Operating income
($ in million) $11,557 (135) (979) (26) — — $10,417
Net income
($ in million) $10,020 (135) (979) (26) (70) 433 $9,243
Shares used in diluted
per share calculation
(millions)
2,494 — — — — — 2,494
Diluted EPS $4.02 — — — — — $3.71
A. Consists of amortization of intangible assets and transaction costs.
B. Stock-based compensation charge was allocated to cost of goods sold, research and development expense, and sales, general and administrative expense.
C. Other represents net losses from non-affiliated investments and interest expense related to amortization of debt discount

-- 59 of 65 --

Reconciliation of Non-GAAP to GAAP Financial Measures (contd.)
Gross Margin Non-GAAP
Acquisition-Related
and Other Costs
(A)
Stock-Based
Compensation
(B)
IP-Related
Costs GAAP
Q3 FY 2023 56.1% (2.0) (0.5) — 53.6%
Q4 FY 2023 66.1% (2.0) (0.5) (0.3) 63.3%
Q1 FY 2024 66.8% (1.7) (0.4) (0.1) 64.6%
Q2 FY 2024 71.2% (0.9) (0.2) — 70.1%
A. Consists of amortization of intangible assets
B. Stock-based compensation charge was allocated to cost of goods sold

-- 60 of 65 --

Reconciliation of Non-GAAP to GAAP Financial Measures (contd.)
Gross Margin
($ in Millions &
Margin Percentage)
Non-GAAP
Acquisition-Related
and Other Costs
(A)
Stock-Based
Compensation
(B)
IP-Related Costs GAAP
FY 2019 $7,233 — (27) (35) $7,171
61.7% — (0.2) (0.3) 61.2%
FY 2020 $6,821 — (39) (14) $6,768
62.5% — (0.4) (0.1) 62.0%
FY 2021 $10,947 (425) (88) (38) $10,396
65.6% (2.6) (0.5) (0.2) 62.3%
FY 2022 $17,969 (344) (141) (9) $17,475
66.8% (1.4) (0.5) — 64.9%
FY 2023 $15,965 (455) (138) (16) $15,356
59.2% (1.7) (0.5) (0.1) 56.9%
YTD Q3 2023 $11,966 (335) (108) — $11,523
57.2% (1.6) (0.5) — 55.1%
YTD Q3 2024 $28,000 (358) (96) (36) $27,510
72.1% (0.9) (0.2) (0.1) 70.9%
A. Consists of amortization of intangible assets and inventory step-up
B. Stock-based compensation charge was allocated to cost of goods sold

-- 61 of 65 --

Operating Income
and Margin
($ in Millions &
Margin Percentage)
Non-GAAP Acquisition
Termination Cost
Acquisition-Related
and Other Costs
(A)
Stock-Based
Compensation
(B)
IP-Related
Costs
Other
(C) GAAP
FY 2019 $4,407 — (2) (557) (35) (9) $3,804
37.6% — — (4.7) (0.3) (0.1) 32.5%
FY 2020 $3,735 — (31) (844) (14) — $2,846
34.2% — (0.3) (7.7) (0.1) — 26.1%
FY 2021 $6,803 — (836) (1,397) (38) — $4,532
40.8% — (5.0) (8.4) (0.2) — 27.2%
FY 2022 $12,690 — (636) (2,004) (9) — $10,041
47.2% — (2.5) (7.4) — — 37.3%
FY 2023 $9,040 (1,353) (674) (2,710) (16) (63) $4,224
33.5% (5.0) (2.5) (10.0) (0.1) (0.2) 15.7%
YTD Q3 2023 $6,816 (1,353) (499) (1,971) — (25) $2,968
32.6% (6.5) (2.4) (9.4) — (0.1) 14.2%
YTD Q3 2024 $22,385 — (446) (2,555) (36) 10 $19,358
57.7% — (1.1) (6.6) (0.1) — 49.9%
Reconciliation of Non-GAAP to GAAP Financial Measures (contd.)
A. Consists of amortization of acquisition-related intangible assets, inventory step-up, transaction costs, compensation charges, and other costs
B. Stock-based compensation charge was allocated to cost of goods sold, research and development expense, and sales, general and administrative expense
C. Comprises of legal settlement cost, contributions, restructuring costs and assets held for sale related adjustments

-- 62 of 65 --

Reconciliation of Non-GAAP to GAAP Financial Measures (contd.)
($ in Millions) Free Cash Flow
Purchases Related to Property
and Equipment
and Intangible Assets
Principal Payments
on Property
and Equipment and Intangible
Assets
Net Cash
Provided by
Operating Activities
FY 2019 $3,143 600 — $3,743
FY 2020 $4,272 489 — $4,761
FY 2021 $4,677 1,128 17 $5,822
FY 2022 $8,049 976 83 $9,108
FY 2023 $3,750 1,833 58 $5,641
YTD Q3 2023 $2,015 1,324 54 $3,393
YTD Q3 2024 $15,732 815 44 $16,591

-- 63 of 65 --

Reconciliation of Non-GAAP to GAAP Financial Measures
($ in Millions) Q4 FY24 Outlook
Non-GAAP gross margin 75.5%
Impact of stock-based compensation expense, acquisition-related costs, and other costs (1.0%)
GAAP gross margin 74.5%
Non-GAAP operating expenses $2,200
Impact of stock-based compensation expense, acquisition-related costs, and other costs 965
GAAP operating expenses $3,165

-- 64 of 65 --



-- 65 of 65 --