Q4 2024 NVIDIA Corp Earnings Call
Good afternoon, My name's, Rob and I'll be your conference operator today at this time I would like to welcome everyone to the Nvidia is fourth quarter earnings call. All lines have been placed on mute to prevent any background noise. After the speakers' remarks, there will be a question and answer session. If you'd like to ask a question. During this time simply press star followed by the number one on your telephone.
Keypad, if you would like to withdraw your question Press Star one again.
Simona Jankowski you may begin your conference.
Simona Jankowski: Thank you good afternoon, everyone and welcome to <unk> conference call for the fourth quarter and fiscal 2024 with me today from Nvidia are Jensen, Huang President and Chief Executive Officer, and Colette, Kress Executive Vice President and Chief Financial Officer.
Speaker Change: I'd like to remind you that our call is being webcast live on Nvidia Investor Relations website. The webcast will be available for replay until the conference call to discuss our financial results for the first quarter of fiscal 2025. The contents of today's call is NVIDIA's property. It can be produced or transcribed without our prior written consent.
Speaker Change: During this call we may make forward looking statements based on current expectations. These are subject to a number of significant risks and uncertainties and our actual results may differ materially for a discussion of factors that could affect our future financial results and business. Please refer to the disclosure in today's earnings release, our most recent forms 10-K and <unk>.
Speaker Change: 10-Q, and the reports that we may file on form 8-K, with the Securities and Exchange Commission.
Speaker Change: All our statements are made as of today February 'twenty, one 'twenty 'twenty four based on information currently available to us.
Speaker Change: Sept as required by law, we assume no obligation to update any such statements. During this call. We will discuss non-GAAP financial measures you can find a reconciliation of these non-GAAP financial measures to GAAP financial measures in our CFO commentary, which is posted on our website with that let me turn the call over to Colette.
Colette M. Kress: Thanks, Simona Q4 was another record quarter revenue of $22 1 billion was up 22% sequentially and up 265% year on year and well above our outlook of 20 billion.
Colette M. Kress: For fiscal 2024 revenue was $60 9 billion and up 126% from the prior year.
Colette M. Kress: Starting with data center data.
Colette M. Kress: Data center revenue for the fiscal 2024 year was four seven.
Colette M. Kress: Seven.
Colette M. Kress: <unk> 5 billion more than tripling from the prior year.
Colette M. Kress: The World has reached a tipping point of new computing.
Colette M. Kress: The trillion dollar installed base of data center infrastructure is rapidly transitioning from general purpose to accelerated computing.
There's more small slows while computing demand continues to skywalk companies may accelerate every workload possible to drive future improvement in performance Tcl and energy efficiency.
Colette M. Kress: At the same time companies have started to build the next generation of modern data centers, what we refer to as a I factories purpose built to refine raw data and produce valuable intelligence in the era of gender today are.
Colette M. Kress: In the fourth quarter data center revenue of $18 4 billion was a record up 27% sequentially and up 409% year on year.
Colette M. Kress: <unk> by the Nvidia Hopper GPU computing platform, along with Infiniband end to end networking.
Colette M. Kress: Compute revenue grew more than five acts and networking Robin you tripled from last year.
Colette M. Kress: We are delighted that supply of Hopper architecture products is improving demand for hopper remains very strong we expect our next generation products to be supply constrained the demand far exceeds supply.
Colette M. Kress: Fourth quarter data center growth was driven by both training and inference of generative AI enlarge language models across a broad set of industries use cases and regions the <unk>.
Colette M. Kress: First fatality in leading performance of our data center platform enables a high return on investment for many use cases, including AI training and inference data processing and a broad range of cuda accelerated workloads, we estimate in the past year approximately 40%.
Colette M. Kress: Our datacenter revenue was for AI inference.
Colette M. Kress: Building and deploying AI solutions has reached virtually every industry many companies across industries, our training and operating their AI models and services at scale.
Enterprises across Nvidia AI infrastructure through cloud providers, including Hyperscale, Gpus specialized and private cloud or on premise.
Colette M. Kress: And videos computing stack extends seamlessly across cloud and on premise environments, allowing customers to deploy with a multi cloud or hybrid cloud strategy.
Colette M. Kress: In the fourth quarter large cloud providers represented more than half of our data center revenue supporting both internal workloads and external public cloud customers.
Colette M. Kress: Microsoft recently noted that more than 50000 organizations use get hub co pilot business to supercharge the productivity of the developers contributing to get hug revenue growth accelerating to 40% year on year.
Colette M. Kress: And co pilot for Microsoft 365 adoption grew faster in its first two months than the two previous major Microsoft 365 Enterprise suite releases Doug.
Colette M. Kress: Consumer Internet companies have been early adopters of AI and represent one of our largest customer categories companies from search to E Commerce, Social media News and video services and entertainment are using AI for deep learning based recommendation systems.
These AI investments are generating a strong return.
Colette M. Kress: Improving customer engagement and conversation and click through rates.
Colette M. Kress: Meta in its latest quarter cited more accurate predictions and improved advertiser performance is contributing to the significant acceleration in its revenue.
Colette M. Kress: In addition, consumer Internet companies are investing in general they are to support content creators advertisers and customers through automation tools for content and AD creation online product descriptions and AI shopping assistance.
Colette M. Kress: Enterprise software companies are applying generative AI to help customers realize productivity gains.
Colette M. Kress: Customers, we've partnered with for both training and inference of generative AI are already seen notable commercial success service now generative AI products in the latest quarter drove their largest ever net new annual contract value contribution of any new product family.
Colette M. Kress: We release.
Colette M. Kress: We are working with many other leading AI and enterprise software platforms, as well, including Adobe data bricks Getty images S. A P.
Colette M. Kress: And snowflake.
Colette M. Kress: The field of foundation large language models is frightening.
Colette M. Kress: Tropic, Google inflection Microsoft Open AI.
Colette M. Kress: And X AI are leading with continued amazing breakthrough and gender today on it.
Colette M. Kress: Exciting companies like adapt AI 'twenty, one character AI cohere Mistral perplexity and runway are building platforms to serve enterprises and creators.
Colette M. Kress: New startups are creating alums to serve the specific languages cultures and customs of the world many regions.
Colette M. Kress: And others are creating foundation models to address entirely different industries like recursion pharmaceuticals and generally.
Take bio medicines for biology. These companies are driving demand for Nvidia AI infrastructure through Hyperscale for GPU specialized cloud providers.
Colette M. Kress: Just this morning, we announced that we collaborated with Google to optimize its state of the art, New Gemma language models to accelerate inference performance on Nvidia Gpus in the cloud data center and PC.
Colette M. Kress: One of the most notable trends over the past year is the significant adoption of AI by enterprises across the industry verticals, such as automotive health care and financial services and.
Colette M. Kress: In video offers multiple application frameworks to help companies adopt AI and vertical domains, such as autonomous driving.
Colette M. Kress: Drug discovery low latencies machine learning for fraud detection or robotics, leveraging our full stack accelerated computing platform.
Colette M. Kress: We estimate the datacenter revenue contribution of the automotive vertical through the cloud or on Prem exceeded $1 billion last year.
Colette M. Kress: Nvidia drive infrastructure solutions include systems and software for the development of autonomous driving including data ingestion creation labeling in AI training plus validation through simulation.
Colette M. Kress: Almost 80 vehicle manufacturers across global Oems, New energy vehicles, trucking robo taxis and tier one suppliers are using Nvidia AI infrastructure to train allo loans and other AI models for automated driving and AI cockpit applications in effect.
Colette M. Kress: Every automotive company working on AI is working with Nvidia.
As Avi algorithms move to video Transformers and more cars are equipped with cameras, we expect in videos automotive data center processing demand to grow significantly.
Colette M. Kress: In health care.
Colette M. Kress: Digital biology, and generative AI are helping to reinvent drug discovery surgery medical imaging and wearable devices, we have built deep domain expertise in healthcare over the past decade, creating the Nvidia Clara healthcare platform and Nvidia.
Colette M. Kress: Fire Nemo degenerative AI service to develop customize and deploy AI Foundation models for computer aided drug discovery.
Colette M. Kress: <unk> Nemo features a growing collection of pre trained bio molecular AI models that can be applied to the <unk> and drug discovery processes.
Colette M. Kress: We announced Rickerson is making available for their proprietary AI model through by Nemo for the drug discovery go system.
Colette M. Kress: In financial services customers are using AI for growing set of use cases from trading and risk management to customer service and fraud detection for.
Colette M. Kress: For example, American express improved fraud detection accuracy by 6% using Nvidia AI.
Colette M. Kress: Shifting to our data center revenue by geography.
Colette M. Kress: Growth was strong across all regions, except for China, where our data center revenue declined significantly following the U S government export control regulations composed in October.
Colette M. Kress: Although we have not received licenses from the U S government to ship restricted products to China. We have started shipping alternatives that don't require a license for the China market.
Colette M. Kress: China represented a mid single digit percentage of our datacenter revenue in Q4.
Colette M. Kress: And we expect it to stay in a similar range in the first quarter.
Colette M. Kress: In regions outside of the U S and China sovereign AI has become an additional demand driver.
Colette M. Kress: <unk> around the world are investing in AI infrastructure to support the building of large language models in their own language on domestic data and in support of the local research and enterprise ecosystem.
Colette M. Kress: From a product perspective, the vast majority of revenue was driven by our Hopper architecture, along with Infiniband networking together they have emerged as the de facto standard for accelerated computing and AI infrastructure.
We are on track to ramp H 200, with initial shipments in the second quarter demand is strong as H 200, nearly doubled the inference performance of each 100.
Colette M. Kress: Networking exceeded a $13 billion annualized revenue run rate our end to end networking solutions defined modern AI data centers, our quantum Infiniband solutions grew more than five X year on year.
Colette M. Kress: Nvidia quantum infiniband as the standard for the highest performance.
Colette M. Kress: I dedicated infrastructures, we are now entering the Ethernet networking space with the launch of our new spectrum ex end to end offering designed for AI optimized networking for the data center.
Colette M. Kress: Spectrum ex introduces new technologies over Ethernet that are purpose built for AI.
Colette M. Kress: Technology is incorporated in our spectrum switch Bluefield, CPU and software stock the Liberal one six X higher networking performance for AI processing compared with traditional Ethernet.
Colette M. Kress: Leading Oems, including Dell, HP, Lenovo and supermicro with their global sales channels are partnering with us to expand our AI solution to enterprises worldwide.
Colette M. Kress: We're on track to ship spectrum X this quarter.
Colette M. Kress: We also made great progress with our software and services offerings, which reached an annualized revenue run rate of $1 billion in Q4.
Colette M. Kress: We announced that Nvidia <unk> cloud will expand its list of partners to include Amazon AWS, joining Microsoft Azure, Google Cloud and Oracle cloud <unk>.
Colette M. Kress: <unk> cloud is used for in videos own AI, R&D and custom model development as well as in video developers. It brings the cuda ecosystem to Nvidia CSP partners.
Colette M. Kress: Okay moving to gaming game.
Colette M. Kress: Gaming revenue was $2 87 billion was flat sequentially and up 56% year on year.
Colette M. Kress: Better than our outlook on solid consumer demand for Nvidia G Force, our TX Gpus during the holidays.
Fiscal year revenue of $10 45 billion was up 15%.
Colette M. Kress: At CES, we announced our G Force RPX 40 Super series family of Gpus.
Colette M. Kress: $599 data lever incredible gaming performance and generative AI capabilities.
Colette M. Kress: <unk> are off to a great start Nvidia AI tensor cores in the Gpus deliver up to 836 AI tops.
Colette M. Kress: Perfect for powering AI for gaming, creating an everyday productivity.
Colette M. Kress: The rich software stock, we offer with our RPX Gpus further accelerates AI with our D. O S. S technologies seven out of eight pixels can be AI generated resulting up to four X faster ray tracing and better image quality and with a tensor RT alone.
Colette M. Kress: For Windows.
Colette M. Kress: Our open source library that accelerates inference performance for the latest large language models generative AI can run up to five X faster on RPX AI Pcs.
Colette M. Kress: At CES, we also announced a wave of new RPX 40 series AI laptops from every major Oems.
Colette M. Kress: These bring high performance gaming and AI capabilities to a wide range of form factors, including 14 inch and thin and light laptops.
Colette M. Kress: With up to 686 top of AI performance. These next generations AIP sees increase agenda today I performance by up to 60 X, making them the best performing.
Colette M. Kress: AI PC platforms.
Colette M. Kress: At CES, we announced Nvidia Avatar cloud engine micro services, which allow developers to integrate state of the art generative AI models into digital avatars.
Colette M. Kress: <unk> won several best of CES 2024 Awards.
Colette M. Kress: Nvidia has an end to end platform for building and deploying generative AI applications for RPX Pcs and workstations. This includes libraries sdk's tools and services developers can incorporate into their journey to AI workloads.
Colette M. Kress: Video is fueling the next wave of general today, our applications coming to the PC with over $100 million RPX Pcs in the installed base and over 500, AI enabled PC applications and games.
Speaker Change: We are on our way.
Speaker Change: Moving to pro visualization.
Speaker Change: Revenue of $463 million was up 11% sequentially and up 105% year on year.
Fiscal year revenue of 1.55 billion was up 1%.
Speaker Change: Sequential growth in the quarter was driven by a rich mix of RPX Ada architecture Gpus continuing to ramp.
Speaker Change: Enterprises are refreshing their workstations to support generative AI related workloads, such as data preparation.
Speaker Change: Fine tuning and retrieval augmented generation.
These key industrial.
Speaker Change: Verticals driving demand include manufacturing automotive and robotics.
Speaker Change: Automotive industry has also been an early adopter of Nvidia omnivorous as it seeks to digitize workflows from design to build stimulate operate and experience their factories in college.
Speaker Change: At CES, we announced that creative partners and developers, including Brooklyn, WTP and zero life are building omni versus powered car configurations.
Speaker Change: Leading automakers like Lotus are adopting the technology to bring new levels of personalization realism and enter activity to the car buying experience.
Speaker Change: Moving to automotive.
Speaker Change: Revenue was $281 million up 8% sequentially and down 4% year on year fiscal.
Speaker Change: Fiscal year revenue of 1.09 billion was up 21% crossing the 1 billion Mark for the first time on continued adoption of the Nvidia drive platform by automakers.
Speaker Change: Nvidia drive Orin is.
Speaker Change: My car computer of choice for software defined a b fleets its.
Speaker Change: His successor and video drive store design for vision Transformers, Ocwen offers more AI performance and integrate a wide range of intelligent capabilities into a single AI compute platforms, including autonomous driving and parking.
Speaker Change: Driver and passenger monitoring and AI cockpit functionality and will be available next year.
Speaker Change: There was solid several automotive customer announcements this quarter.
Speaker Change: The auto great wall Motor Zika.
Speaker Change: Premium EV subsidiary of DNA and Xiaomi equally.
<unk> announced new vehicles built on Nvidia.
Speaker Change: Moving to the rest of the P&L.
Speaker Change: GAAP gross margins expanded sequentially to 76% and non-GAAP gross margins to 76, 7% on strong data center growth and mix.
Speaker Change: Our gross margins in Q4 benefited from favorable component cost.
Speaker Change: Sequentially GAAP operating expenses were up 6% and non-GAAP operating expenses were up 9%, primarily reflecting higher compute and infrastructure investments and employee growth.
Speaker Change: In Q4, we returned $2 8 billion to shareholders in the form of share repurchases and cash dividends during fiscal year 'twenty four we utilized cash of 9.9 billion towards shareholder returns, including $9 5 billion in share repurchases.
Speaker Change: Let me turn to the outlook for the first quarter.
Speaker Change: Total revenue is expected to be 24 billion plus or minus 2%.
Speaker Change: We expect sequential growth in datacenter and probus, partially offset by seasonal decline in gaming.
Speaker Change: GAAP and non-GAAP gross margins are expected to be 76, 3% and 77%, respectively, plus or minus 50 basis points.
Speaker Change: Similar to Q4 Q1 gross margins are bad is benefiting from favorable component cost.
Speaker Change: Beyond Q1 for the remainder of the year, we expect gross margins to return to the mid seventies per cent range.
Speaker Change: GAAP and non-GAAP operating expenses are expected to be approximately $3 5 billion and $2 5 billion respectively.
Speaker Change: School year, 2025, GAAP and non-GAAP operating expenses are expected to grow in the mid 30% range as we continue to invest in the large opportunities ahead of us.
Speaker Change: GAAP and non-GAAP other income and expenses are expected to be an income of approximately $250 million, excluding gains and losses from non affiliated investments.
Speaker Change: GAAP and non-GAAP tax rates are expected to be 17% plus or minus 1%, excluding any discrete items.
Speaker Change: Further financial details are included in the CFO commentary and other information available on our IR website.
Speaker Change: In closing, let me highlight some upcoming events for the financial community, we will attend the Morgan Stanley Technology, and media and Telecom conference in San Francisco on March four.
Speaker Change: And the TD Cohen's 44th annual Health Care Conference in Boston on March 10th.
Speaker Change: And of course George.
Speaker Change: US for our annual GTC Conference, starting Monday March 18th in San Jose, California.
We held in person for the first time in five years DTC will kick off with Jenson keynote and we will host a Q&A session for financial analysts. The next day March 16th.
Speaker Change: Time, we are now open the call for questions. Operator would you please poll for questions.
At this time I would like to remind everyone in order to ask a question Press Star then the number one on your telephone keypad well pause for just a moment to compile the Q&A roster. As a reminder, please limit yourself to one question.
Your first question comes from the line of <unk> Hari from Goldman Sachs. Your line is open.
Hari: Hi, Thank you so much for taking the question and congratulations on the really strong results Mike.
Hari: My question is for Jensen on the data center business.
Hari: Clearly youre doing extremely well.
Hari: The business I'm curious, how your expectations for calendar 'twenty four and.
Hari: 25.
Jensen Huang: <unk> over the past 90 days and as you answer the question I was hoping you can touch on some of the newer buckets within data center things like software.
Jensen Huang: Sovereign AI I think you've been pretty vocal about how to think about that medium to long term and recently there was an article about video potentially participating in the ASIC market is there any credence to that and if so how should we think about you guys playing in that market over the next.
Speaker Change: Thank you.
Speaker Change: Thanks Toshi.
Speaker Change: Hi.
Speaker Change: Let's see.
Speaker Change: There were three questions one more time the first question.
Speaker Change: Can you <unk>.
Speaker Change: Well I guess your expectations for data center.
Speaker Change: Thank you.
Speaker Change: Okay, Yes.
Speaker Change: Well you know we guide one quarter at a time.
Speaker Change: But fundamentally the conditions.
Speaker Change: Our excellent for continued growth calendar 2014 calendar 'twenty, five and beyond and let me tell you why.
Speaker Change: We're at the beginning of two.
Speaker Change: Industry wide.
Speaker Change: Transitions.
Speaker Change: And both of them are industry wide. The first one is a transition from general to accelerated computing.
Speaker Change: General purpose computing as you know is starting to run out of steam and you could tell by the csp's extending and many data centers, including our own for general purpose computing extending the depreciation from four to six years. There's just no reason to update with more Cpus, when you can't fundamentally and dramatically enhance.
Speaker Change: It's throughput like you used to.
Speaker Change: And so you have to accelerate everything this is what <unk> been pioneering for some time.
And with accelerated computing, you could dramatically improve your energy efficiency, you can dramatically improve your cost.
Speaker Change: In data processing by 20 to one.
Speaker Change: Huge numbers.
Speaker Change: And of course the speed.
Speaker Change: That speed is so incredible.
Speaker Change: We enabled a second industry wide transition called generative AI.
Speaker Change: Jeremy.
Jeremy: Sure what's been talked plenty plenty about it during the call, but remember generative AI is a new application.
Jeremy: This is enabling a new way of doing software new types of software being created.
Jeremy: It is a new way of computing.
Jeremy: You can't do generative AI on traditional general purpose computing you have two accelerated.
Jeremy: And the third is it is enabling a whole new industry.
Jeremy: And this is this is something worthwhile to take a step back and look at it.
Jeremy: It connects to your last question about sovereign AI.
Jeremy: A whole new industry in the sense that for the very first time, a data center is not just about <unk>.
Jeremy: Computing data and storing data in serving the employees of the company.
Jeremy: We now have a new type of datacenter.
Jeremy: That is about <unk>.
Jeremy: AI.
Jeremy: Generation and.
And AI generation factory.
Jeremy: And you've heard me describe it as AI factories.
Jeremy: But basically it takes raw material, which is data.
Jeremy: It transforms it with these AI supercomputers that Nvidia built.
Jeremy: And it turns them into incredibly valuable.
Tokens.
Jeremy: These tokens are what people experience on the amazing amazing chat GPT or mid journey or.
Jeremy: Hi.
Search these days are augmented by that all of your recommends systems are now augmented by that the hyper personalization that goes along with it.
Jeremy: All of these incredible startups and digital biology.
Jeremy: Generating proteins and generating chemicals.
Jeremy: The list goes on and so all of these tokens are generated.
Jeremy: In a very specialized type of datacenter and this data center, we call it.
Jeremy: Supercomputers in AI generation factories.
Jeremy: But we're seeing diversity.
Jeremy: One of the other.
Jeremy: So at the foundation is that the way it manifests in two new markets.
Jeremy: <unk> is an all of the diversity that youre seeing us in.
Jeremy: One the amount of influence that we do.
Jeremy: It is just off the charts now almost every single time, you interact with chat GPT you'd know that we're interesting every time you use in the journey, where inferencing every time you see amazing these sort of videos that are being generated or runway to videos that their editing Firefly Nvidia doing inferencing.
The inference part of our business has grown tremendously we estimate about 40%.
Jeremy: The amount of training is continuing because these models are getting larger and larger the amount of influence is increasing but we're also diversifying into.
Jeremy: New industries.
Jeremy: The large CSP.
Jeremy: We are still continuing to build out you could see from their capex and their discussions.
Jeremy: But theres a whole new category called GPU specialized.
Jeremy: <unk>.
Jeremy: They specialize in video Nvidia AI infrastructure Gpus specialized csp's.
Jeremy: <unk> enterprise software platforms.
Jeremy: Deploying AI service now is just a really really great example.
Jeremy: Adobe Theres see others, SAP and others you.
Jeremy: We see consumer Internet services that are now augmenting all of their services with the past with generative AI. So they can have even more hyper personalized.
Jeremy: Content to be created.
Jeremy: You see us talking about industrial.
Jeremy: Today, our now our industries represent.
Jeremy: Multibillion dollar businesses auto health financial services.
Jeremy: In total our vertical industries are multi multibillion dollar businesses now and of course sovereign AI the reason for sovereign AI.
Jeremy: It has to do with the fact that.
Jeremy: The language.
Jeremy: The knowledge the history the culture of each region are different.
And they own their own data they would like to use their data.
Jeremy: Trade it with to create their own digital intelligence and provisioning to harness that raw materials themselves. It belongs to them each one of the regions around the world the data belongs to them.
Jeremy: The data is most useful to their society and so they want it.
Jeremy: Protect the data they want to transform it themselves value added transformation into into AI and provision those services themselves.
Jeremy: So we're seeing sovereign AI.
Jeremy: Infrastructure is being built in.
Jeremy: In Japan.
Jeremy: And <unk>.
Jeremy: Canada and France.
Jeremy: So many other regions.
Jeremy: And so my expectation is that what is being experienced here in the United States and the west.
Jeremy: We will surely be replicated around the world and these AI generation factories are going to be in every industry every company every region.
Jeremy: And so so I think the last the last.
Jeremy: This last year, we've seen.
Jeremy: <unk>, AI really becoming a whole new <unk>.
Jeremy: Implication space, a whole new way of doing computing, a whole new industry is being formed.
Jeremy: And that's driving our growth.
Jeremy: Your next question comes from the line of Joe Moore from Morgan Stanley. Your line is open.
Joe Moore: Great. Thank you.
Joe Moore: And to follow up on the 40% of revenue is coming from entrance.
Joe Moore: That's a bigger number than I expected can you give us some sense of where that number was maybe a year before how much youre seeing growth around llm's from and brands.
Joe Moore: And how are you measuring that is that I assume it's in some cases the same gpus you used for training and inference.
Joe Moore: How solid is that measurement. Thank you.
Speaker Change: I'll go backwards.
Speaker Change: The estimate is probably understated.
Speaker Change: But we estimated it and let me tell you why.
Speaker Change: Hi.
Speaker Change: Whenever a year ago a year ago.
Speaker Change: The recommended systems that people are.
Speaker Change: When you run the Internet the news the videos.
Speaker Change: The music.
Speaker Change: The products that are being recommended to you because as you know the Internet is trillions I don't know how many trillions, but.
Speaker Change: Trillions of things out there and your phone is.
Speaker Change: Three inches square.
Speaker Change: And so the ability for them to fit all of that information down to something such a small real estate is through a system, an amazing system called recommend or systems.
Speaker Change: These recommend or systems.
Speaker Change: Used to be.
Speaker Change: All based on CPU CPU.
Speaker Change: Approaches.
Speaker Change: But the recent migration to deep learning.
Speaker Change: And now generative AI has really put these recommended systems now.
Speaker Change: Directly into the path of GPU acceleration needs GPU acceleration for the embedding in these GPU acceleration for the nearest neighbors search it needs GPU accelerating for re ranking and needs GPU acceleration to generate.
Speaker Change: The augmented information for you. So gpus are in every single step of our recommended system now and as you know recommend their system is the single.
Speaker Change: Largest software engine on the planet.
Speaker Change: Most every major major company in the World has to run these large recommended systems, but.
Speaker Change: Whenever users chat GPT has been inference whenever you.
Speaker Change: Here about mid journey into just a number of things that they're generating four.
Speaker Change: For consumers when you when you see.
Speaker Change: Getty the work that we do with Getty and Firefly from Adobe. These are all generative models.
Speaker Change: The list goes on and and none of these items as I mentioned that existed a year ago a 100%.
Speaker Change: Your next question comes from the line of Stacy <unk> from Bernstein Research. Your line is open.
Stacy: Hi, guys. Thanks for taking my question.
Stacy: I wanted to call out I wanted to touch on your comment that you expected. The next generation of products I feel that Ed Blackwell to be supply constrained.
Stacy: Can you dig into that a little bit.
Stacy: What is the driver of that wire why does that get constrained as hopper is easing up and how long do you expect that to be constrained like do you expect.
Stacy: Exploration to be constrained like like all the way through calendar 'twenty five like when do those start to ease.
Speaker Change: Yeah. The first thing is.
Speaker Change: Overall, our supply is improving.
Speaker Change: Overall.
Speaker Change: Our supply chain is just doing an incredible job for us.
Speaker Change: Everything from of course, the wafer is the packaging the memories.
Speaker Change: All of the power regulators.
Speaker Change: Two two.
Speaker Change: Transceivers and.
Speaker Change: Hi.
Speaker Change: Networking and cables and.
Speaker Change: You name it the list of components that we ship as you know people think that Nvidia Gpus is like a chip, but the Nvidia Hopper GPU is 35000 parts.
Speaker Change: It weighs 70 pounds.
Speaker Change: These things are really complicated things, we built people call. It AI supercomputer for good reason.
Speaker Change: Have you ever looking look in the back of the data center.
Speaker Change: The systems the cabling system is mindboggling. It is the most <unk>.
Speaker Change: <unk> complex cabling system for networking the world's ever seen.
Speaker Change: Our infiniband business grew <unk> year over year, the supply chain is really doing fantastic supporting us.
Speaker Change: And so so overall.
Speaker Change: Supply is improving.
Speaker Change: We expect that demand will continue to be stronger.
Speaker Change: Then our supply provides and and.
Speaker Change: Through the year and we'll do our best.
Speaker Change: The cycle times are improving.
Speaker Change: And we're going to continue to do our best however.
Speaker Change: Whenever we have new products as you know it ramps from zero.
Speaker Change: Two a very large number and you can't do that overnight.
Everything everything is ramped up.
Speaker Change: Step up.
Speaker Change: And so.
Speaker Change: Whenever we have a new generation of products and right now we are ramping.
Speaker Change: H two hundreds.
Speaker Change: There is no way, we can reasonably kept keep up.
Speaker Change: On demand in the short term as we ramp.
Speaker Change: We're ramping spectrum X.
Speaker Change: We're doing incredibly well with spectrum X, it's our brand new product.
Into the world of Ethernet.
Speaker Change: Infiniband is the standard for <unk>.
Speaker Change: AI dedicated.
Speaker Change: Systems.
Speaker Change: Ethernet with spectrum ex Ethernet is just not a very good scale out system.
Speaker Change: But with spectrum X weave augmented layered on top of Ethernet.
Speaker Change: Fundamental new capabilities like adaptive routing congestion control noise.
Speaker Change: Noise isolation, where traffic isolation.
Speaker Change: So that we could.
Speaker Change: Optimize Ethernet for AI, and so infiniband will be our AI dedicated infrastructure spectrum X will be our AI optimized.
Speaker Change: Networking and that is ramping and so so well.
Speaker Change: With all of new products.
Demand is greater than supply and Thats, just kind of the nature of new products.
Speaker Change: We worked as fast as we can.
Speaker Change: To catch up with the demand, but overall overall net net overall our supply is increasing.
Very nicely.
Speaker Change: Your next question comes from the line of Matt Ramsay from TD Cowen Your line is open.
Matt Ramsay: Good afternoon Jensen Colette congrats on the results.
Matt Ramsay: I wanted to ask I guess, a two part question and it comes at what safety was just getting at on your demand being significantly more than than your supply, even though supply is improving and I guess.
Matt Ramsay: The two sides of the question or I guess first for Colette.
Colette M. Kress: How are you guys thinking about allocation of product.
Colette M. Kress: In terms of customer readiness to deploy.
Colette M. Kress: And sort of monitoring.
Colette M. Kress: Any kind of buildup.
Colette M. Kress: Product that might not yet be turned on and then I guess Jensen for you I'd be really interested to hear you speak a bit about.
Colette M. Kress: A thought that you and your company are putting into the allocation of your product.
Colette M. Kress: Across customers, many of which compete with each other.
Colette M. Kress: Across industries too.
Colette M. Kress: It's a smaller startup company to things in the health care Arena to government.
Jensen: It's a very very unique technology that you are enabling and I'd be really interested to hear you speak a bit about how you think about quote unquote fairly allocating sort of for the good of your company, but also for the good of the industry.
Speaker Change: Let me first start with your.
Speaker Change: Question thinks about.
How we are working with our customers as they look into how they are building out.
Speaker Change: Their GPU instances.
Speaker Change: And our allocation process.
Speaker Change: Folks that we work with our customers that we work with I have been partners with us for many years.
Speaker Change: As we have been assisting them both in what they set up in the cloud as.
Speaker Change: One is they are setting up internally many of these providers have multiple products going at one time to serve so many different needs across their end customers, but also what they need internally.
Speaker Change: So they are working in advance of course thinking about those new clusters that they will need and our discussions with them continue not only on our hopper architecture, but helping them understand the next wave and getting their interest in getting.
Speaker Change: Their outlook for the demand that they want so it's always a moving process in terms of.
Speaker Change: What they will purchase what is still being built and what is in use for end customers of the relationships that we built and their understanding of the sophistication of the build is really helped us with that allocation and both helped us with our communications with them.
Speaker Change: First our csp's.
Speaker Change: Have a very clear view.
Speaker Change: Our <unk>.
Speaker Change: Product roadmap and transitions.
Speaker Change: And.
Speaker Change: That transparency with our CSP.
Speaker Change: It gives them the.
Speaker Change: Our confidence.
Speaker Change: Of.
Speaker Change: Which products to place and where and when.
Speaker Change: And so.
Speaker Change: They know there they know.
Speaker Change: They know the timing to the best of our ability and they know quantities.
And of course allocation.
Speaker Change: We allocate fairly.
Speaker Change: We allocate fairly due.
Speaker Change: To the best of our best we can to allocate fairly.
Speaker Change: And to avoid allocating unnecessarily.
Speaker Change: As you mentioned earlier why allocate something when a data center is not ready.
Speaker Change: Nothing nothing is.
Speaker Change: More difficult then to have anything sit around and so so allocate fairly and to avoid allocating unnecessarily and where we do well.
Speaker Change: The question that you asked about the end markets you know that we have.
Speaker Change: An excellent ecosystem.
Speaker Change: With.
Oems Odm's, Csp's, and very importantly, and markets.
Speaker Change: What Nvidia is really unique about is are we bring our customers, we bring our partners <unk> and Oems, we bring them customers.
Speaker Change: The biology companies to health care companies financial services companies.
AI developers large language model developers.
Speaker Change: Animas vehicle companies robotics companies or as just a giant suite of robotics companies that are emerging.
Speaker Change: Their warehouse robotics to surgical robotics to humanoid robotics, all kinds of really interesting robotics companies agricultural robotics companies.
All of these startups.
Speaker Change: Large companies health care financial services, and auto and such.
Speaker Change: Our working on Nvidia platform, we support them directly.
Speaker Change: And oftentimes, we can have a twofer.
Speaker Change: By allocating towards CSP, and bringing the customer to the CSP at the same time.
And so.
Speaker Change: So this ecosystem.
Speaker Change: Youre, absolutely right that its vibrant.
But at the core of it we want to allocate fairly.
Speaker Change:
Speaker Change: Avoiding waste and.
Speaker Change: And looking for opportunities to connect partners and end users.
Speaker Change: We're looking for those opportunities all the time.
Speaker Change: Your next question comes from the line of Timothy Arcuri from UBS. Your line is open.
Timothy Michael Arcuri: Thanks, a lot I wanted to ask about how you are converting backlog into revenue obviously lead times for your products have come down quite a bit.
Timothy Michael Arcuri: You didn't talk about the inventory purchase commitments, but if I sort of add up your inventory plus the purchase commitment and your prepaid support supply sort of the aggregate of your supply it was actually down a touch.
Timothy Michael Arcuri: How should we read that is that just you, saying that you don't need to make as much of a financial commitment to your suppliers because the lead times are lower or is that maybe you are reaching some sort of steady state. We are closer to filling your photo book in your backlog.
Timothy Michael Arcuri: Yes, So let me let me highlight on those three different areas of how we look at our suppliers Youre correct. Our inventory on hand, given our allocation that we're on we're trying to as things come into inventory immediately work to ship them to our customers I think our customer appreciates.
Timothy Michael Arcuri: Our ability to meet the schedules that we've looked for the second piece of it is our purchase commitments our purchase commitments have many different components into it.
Timothy Michael Arcuri: Components that we need for manufacturing, but also often we are procuring capacity that remained at the length of that need for capacity or the length of the components are all different some of them may be for the next two quarters. Some of them may be for multiple years I can say the same regarding our prepaid.
Timothy Michael Arcuri: Our prepays are pre designed to make sure that we have the reserve capacity that we need at several of our manufacturing suppliers as we work forward. So wouldnt read into anything regarding approximately about the same numbers as we are increasing our supply all of them just have different links.
Timothy Michael Arcuri: As we have sometimes hard to buy things in long lead times or things that need a capacity build for us.
Timothy Michael Arcuri: Yeah.
Timothy Michael Arcuri: Yes.
Timothy Michael Arcuri: Your next question comes from the line of Ben Rights from Melius Research. Your line is open.
Ben Rights: Yes. Thanks.
Ben Rights: Congratulations on the results I wanted to talk about your comment regarding gross margins.
Ben Rights: <unk>.
Ben Rights: They should go back to the mid seventies.
Ben Rights: If you don't mind unpacking that.
Ben Rights: <unk> also.
Ben Rights: Is that to date the HBM.
Ben Rights: Content and unique products.
Ben Rights: <unk>.
Ben Rights: What do you think.
Ben Rights: What are the drivers of that comment thanks, so much.
Speaker Change: Yeah. Thanks for the question we highlighted.
Speaker Change: In our opening remarks.
Speaker Change: Really about our Q4 results and our outlook for Q1, both of those quarters are unique.
Speaker Change: Those two quarters are unique in their gross margin as they include some benefit from favorable.
Speaker Change: Component cost in.
Speaker Change: In the supply chain kind of across both our compute and networking and also in several different stages of our manufacturing process.
Speaker Change: Looking forward, we have visibility into a mid seventies gross margin for the rest of the fiscal year, taking us back to where we were before this Q4 and Q1.
Speaker Change: Peak that we've had here.
So we're really looking at just a balance of our mix mix is always going to be our largest driver of what we will be shipping.
Speaker Change: For the rest of the year and those are really just the drivers.
Speaker Change: Your next question comes from the line of C. J Muse from Cantor Fitzgerald. Your line is open yes.
Yes, good afternoon, and thank you for taking the question.
Picture question Free Jensen, when you think about the $1 million improvement in GPU compute over the last decade, and the expectations for similar improvements in the next how do your customers think about the long term usability of their Nvidia investments that we're making today today's training clusters become tomorrow's entrance clusters, how do you see this playing out thank you.
Speaker Change: Hey, C J thanks for the question.
Jensen: Yes, that's the really cool part.
Speaker Change: You look at look at the reason why we're able to.
CJ: To improve performance so much is because.
CJ: We have two characteristics about our platform one.
CJ: Is that has accelerated.
CJ: And two it's programmable.
CJ: It's not brittle NV.
CJ: And video is the only architecture that has gone from the very very beginning literally the very beginning when C N N and Alex Kruszewski in Elliot Susquehanna and Jeff Hinton first revealed Alex net.
CJ: All the way through R&M to Lstm's to every <unk> deep learning <unk> two transformers to every single version every single version of every species that have come along vision Transformers multi modality transformers that every single and an hour time sequenced often and every single variable.
CJ: Variation every single species of AI.
CJ: That has come along we've been able to.
Supported opt.
CJ: Optimize our stack for it and deploy it into our installed base.
CJ: This is really the great amazing part on the one hand.
CJ: Can invent new architectures and new technologies like.
CJ: Our tensor cores like our transformer engine for tensor cores.
CJ: Improved new numerical formats and structures of processing like we've done with the different generations of tensor cores. Meanwhile, supporting the installed base at the same time and so.
CJ: As a result, we take all of our new software algorithm invest.
CJ: Inventions.
CJ: All of the inventions, new inventions of models of the industry and it runs on our installed base on the one hand on the other hand, whenever we see something revolutionary.
CJ: Like Transformers, we can create something brand new like the Hopper transformer engine and implemented into future and so we simultaneously have this ability to.
CJ: Bring software to the installed base, and keep making a better and better and better. So our customers installed base is enriched over time with our new software on the other hand for new technologies create revolutionary capabilities.
CJ: Don't be surprised if in our future generation all of a sudden amazing breakthroughs.
CJ: In large language models were.
CJ: Possible and those breakthroughs.
CJ: Some of which will be in software because they run cuda will be made available to the installed base and so we carry everybody with us on the one hand, we made giant breakthroughs on the other hand.
CJ: Your next question comes from the line of Aaron Rakers from Wells Fargo. Your line is open.
Aaron Rakers: Yes, thanks for taking the question I wanted to ask about the China business I know that.
Aaron Rakers: In your prepared comments, you said that you started shipping some alternative solutions into China, you also put.
Aaron Rakers: Pointed out that you expect that contribution to continue to be about mid single digits.
Speaker Change: I know youre towards data center business. So I guess the question is what is the extent of products that you're shipping today into the China market.
Speaker Change: And why should we not expect that maybe other alternative solutions come to the market and expand your breath to participate in that in that opportunity again. Thank you.
Speaker Change: Okay.
Speaker Change: At the core remember the U S government want to limit.
Speaker Change: The latest capabilities.
Speaker Change: <unk>.
Speaker Change: Nvidia accelerated computing and AI.
Speaker Change: But to the Chinese market.
Speaker Change: And.
Speaker Change: The U S government, we'd like to see us be successful in China as possible.
Speaker Change: Within those two constraints.
Speaker Change: Within those two pillars, if you will.
Speaker Change: The restrictions.
Speaker Change: And so we had to pause.
Speaker Change: When the new restrictions came out.
Speaker Change: We immediately passed.
So we understood what the restrictions are.
Speaker Change: Re.
Speaker Change: Configured our products.
Speaker Change: In a way that is not software.
Speaker Change: Packable in any way.
Speaker Change: And that took some time.
Speaker Change: So we reset we reset our product offering to China and now we're sampling.
Speaker Change: Two customers in China, and we're going to do our best.
Speaker Change: To compete in a marketplace and succeeded in a marketplace.
Speaker Change: Within the within the specifications of the restriction and so that's it.
Speaker Change: We this last quarter.
Speaker Change: Our business significantly declined.
Speaker Change: As we as we paused.
Speaker Change: In the marketplace, we stopped shipping in the marketplace. We expect this quarter to be about the same.
But after that hopefully we.
Speaker Change: We can go compete for our business.
Speaker Change: And do our best and we'll see how it turns out.
Speaker Change: Okay.
Speaker Change: Your next question comes from the line of harsh Kumar from Piper Sandler Your line is open.
Harsh Kumar: Yeah, Hey, Jensen Colette and video team first of all congratulations on a stunning quarter and guide I wanted to talk about.
Harsh Kumar: But the budget software business and it's pleasing to hear that it's over $1 billion, but I was hoping Jensen Colette. If you could just help us understand what the different parts and pieces are for the software business in other words, just help us unpack it a little bit. So we can get a better understanding of where that growth is coming from.
Colette M. Kress: Let me take a step back and explain the fundamental reason why.
Speaker Change: And video will be very successful software.
Speaker Change: So first as you know.
Speaker Change: Accelerated computing really grew.
Colette M. Kress: In the cloud in the cloud the cloud service providers have really large engineering teams and we've worked with them.
Colette M. Kress: In a way that allows them to operate and manage their own business.
Colette M. Kress: And whenever there are any issues, we have large teams assigned to them.
Colette M. Kress: And their engineering teams are working directly with our engineering teams and we.
Colette M. Kress: Enhance we fix we maintained we patch.
Colette M. Kress: The complicated stack of software.
Colette M. Kress: Thats involved in accelerated computing as you know accelerated computing is very different than general purpose computing youre not.
Colette M. Kress: Starting from a.
Colette M. Kress: A program like C plus plus you can pilot and things run on all your Cpus.
<unk> stacks of software necessary for every domain from data processing sequel versus sequels structured data versus versus all the images.
Colette M. Kress: Taxed in PDF, which is unstructured.
Two.
Colette M. Kress: Classical machine learning to computer vision to speech.
Colette M. Kress: <unk> large language models.
Colette M. Kress: All recommended systems all of these things require different <unk>.
Colette M. Kress: Software stacks, that's the reason why Nvidia has hundreds of libraries.
Colette M. Kress: If you don't have software you can't open new markets. If you don't have software you can't open and enable new applications software is fundamentally necessary for accelerated computing. This is the fundamental difference between accelerated computing in general purpose computing that most people took a long time to understand and now people under.
Colette M. Kress: <unk> that the software is really key in the way that we work with Csp's, that's really easy.
Colette M. Kress: We have large teams are working with are large teams. However.
Colette M. Kress: Now that generative AI is enabling every enterprise and every enterprise software company.
Colette M. Kress: To embrace <unk>.
Colette M. Kress: Accelerated computing.
Colette M. Kress: And when it is now essential to embrace accelerated computing because.
Colette M. Kress: It is no longer possible.
Colette M. Kress: So no longer likely anyhow to sustain.
Colette M. Kress: Improved throughput through just general purpose computing.
Colette M. Kress: All of these enterprise software companies and enterprise companies don't have large engineering teams to be able to maintain and optimize.
Colette M. Kress: Their software stack to run across all of the world's clouds and private clouds and on Prem. So we are going to do the management the optimization.
<unk>.
Colette M. Kress: Patching the tuning.
Colette M. Kress: The installed base optimization.
Colette M. Kress: For all of their software stacks and we contain horizon.
Into our stack, we call Nvidia AI enterprise.
Colette M. Kress: And the way we go to market with it is.
Colette M. Kress: Think of that Nvidia AI enterprise now as a run time like an operating system, it's an operating system for artificial intelligence.
Colette M. Kress: And we charge $4500 per GPU per year.
Colette M. Kress: And.
Colette M. Kress: My guess is that that.
Colette M. Kress: Every enterprise in the world.
Colette M. Kress: Every software enterprise company that are deploying software and all the clouds private clouds and on Prem.
Colette M. Kress: We'll run on Nvidia AI enterprise, especially obviously for our Gpus and so this is going to.
Colette M. Kress: Likely be a very significant business over time.
Colette M. Kress: We're off to a great start.
Colette M. Kress: And.
Colette M. Kress: Hi.
Colette M. Kress: Colette mentioned that it's already at $1 billion run rate and Oh.
Colette M. Kress: Just getting started.
Colette M. Kress: Thank you I will now turn the call back over to Jensen Huang CEO for closing remarks.
Jensen Huang: The computer industry is making two simultaneous platform shifts at the same time.
Jensen Huang: The trillion dollar install base of data centers.
Jensen Huang: Is transitioning from general purpose to accelerated computing.
Jensen Huang: Every data center will be accelerated.
Jensen Huang: So the world can keep up with the computing demand with increasing throughput, while managing costs and energy.
Jensen Huang: The incredible speed up between video enabled.
Jensen Huang: The Navy.
Jensen Huang: In video enabled a whole new computing paradigm generative AI.
Jensen Huang: Where software can learn.
Jensen Huang: Understand.
Jensen Huang: And generate any information from human language.
Jensen Huang: To the structure of biology, and <unk> World.
Jensen Huang: We are now at the beginning of a new industry.
Jensen Huang: <unk> AI dedicated data centers process massive raw data to refine it into digital intelligence.
Jensen Huang: Like AC power generation plants of the last industrial Revolution.
Jensen Huang: AI supercomputers are essentially.
Jensen Huang: Generation factories of this industrial Revolution.
Jensen Huang: Every company and every industry is fundamentally built on their proprietary business intelligence.
Jensen Huang: And in the future.
Jensen Huang: Their proprietary generative AI.
Jensen Huang: Generative AI has kicked off a whole new investment cycle to build the next trillion dollars of infrastructure of AI generation factories.
Jensen Huang: We believe these two trends will drive a doubling of the world's datacenter infrastructure installed base in the next five years and will represent an annual market opportunity in the hundreds of billions.
Jensen Huang: This new AI infrastructure will open up a whole new world of applications not possible today.
We started the AI journey with the Hyperscale cloud providers and.
Jensen Huang: And consumer Internet companies.
Jensen Huang: And now every industry is on board from automotive.
Jensen Huang: To healthcare.
Jensen Huang: Financial services.
Jensen Huang: To industrial to.
Jensen Huang: To telecom.
Jensen Huang: Media and entertainment.
Jensen Huang: NVIDIA's full stack computing platform.
Jensen Huang: With industry specific applications frameworks, and a huge developer and partner ecosystem gives.
Jensen Huang: It gives us the speed.
Jensen Huang: Scale and reach to help every company to help companies in every industry become an AI company.
Speaker Change: We have so much to share with you at next month's GTC in San Jose.
Speaker Change: So be sure to join us.
Speaker Change: We look forward to updating you on our progress next quarter.
Speaker Change: This concludes today's conference call you may now disconnect.
Speaker Change: Okay.
Speaker Change: Okay.
Speaker Change: Yeah.
Speaker Change: Okay.