NVIDIA (NVDA) 16 Aug 192020 Q2 Earnings call transcript

Good afternoon. My name is Christina, and I will be your conference operator today. Welcome to NVIDIA’s financial results conference call. [Operator Instructions] I will now turn the call over to Simona Jankowski from Investor Relations to begin your conference.

Thank you. welcome fiscal conference XXXX. and second for everyone of quarter afternoon, to call NVIDIA’s the Good and Vice With NVIDIA President Officer are on Jen-Hsun Kress, me the Huang, and call Chief Financial and Chief President today from Colette Officer. Executive Executive

to replay like our that XXXX. call financial NVIDIA’s discuss until on the you webcast webcast will results call website. our third fiscal Investor conference without The Relations be remind today’s The being quarter our for be live call would consent. It I property. written is NVIDIA’s reproduced of available of the transcribed for or content prior to is can’t

and materially. we call, expectations. subject number this uncertainties, These During actual are risks a and may results make on our forward-looking differ statements based to current significant may of

the a the recent factors of and release, please refer future could today’s obligation statements. XX-K with the our required Form as our file and All based that our business, XXXX, XX, affect available Exchange earnings discussion Securities For in and any update assume information Commission. made currently we disclosure August today, may reports to we no law, us. on on Except most Form financial results of by to that XX-Q statements as to such are and X-K

During this call, we will discuss non-GAAP financial measures.

our GAAP a You on non-GAAP in measures financial to CFO of our can which measures reconciliation posted website. find financial commentary, is these

me let turn over Colette. that, With call to the

XX% XX% of $X.XX with year-on-year gaming revenue Thanks, in was with sequentially. and down Simona. line XX% outlook, up our revenue XX% year-on-year our sequentially. QX down Starting billion, up was business, and billion $X.XX

and SUPER up growth the strong in of laptops our we quarter gaming launched ever number gamers, our wrapped our We new laptops studio desktop for launched for RTX the are creators. pleased lineup when greatest sequential with RTX

and RTX Call real-time gamers a the SUPER year in XX% we GPUs and industry next-generation has high strengthens delighting best get announced is the graphics. computer blockbuster Cyberpunk to taking tracing the for end Dogs: look X leadership unveiled seasons. our the from quickly as in XXXX, best-in-class of market for performance tracing and current performance the our to and boost of Warfare, into lineup We number efficiency A holiday with of great. SUPER Duty: school have by Wolfenstein: The tracing ray titles have and to GPUs, to both ray performance NVIDIA storm Turing forward AAA delivered been earlier. era up modern tracing, Ray define launched and gaming Watch Legion come GPUs a ray In power the July, support delivering the of Modern response initial games. the we GeForce of including gaming growing and RTX back shopping Youngblood. These

speed crafted core. and up a around graphic and Max-Q a the laptop that form our architecture Excitement the enables most They factors Turing ramping thin performance The of with next-generation to GeForce the are can gaming Xx models ray one without be energy-efficient light a over driver ray gaming Cyberpunk these holiday tracing. only to in RTX or business The support our called for performance cards hardware the ahead deliver the to XXX-plus as NVIDIA market deliver titles and the of continues growth console. record of anticipated are laptop GameSpot dedicated high-end back standout of combination of school technology desktop decade. OEMs beautifully season. is X tremendous. tracing a GPUs games

ultra the and Pro. not professional-grade deliver to RTX image traditionally Nintendo been YouTube, design such NVIDIA of market models RTX performance have are from by of XX total freelancers photorealistic have as workstations the creators retail extends of of creators. the their workflows optimized video. our a large, production enhancement laptops, by At artist Studio Studio that new In Computex Studio are RTX ray to underserved major than to ramp RTX online of age laptops Sequential the GPUs RTX high-resolution and in benefited platform the console. growth gaming our meet and Switch and A unveiled through models access complex software, that’s May, we designed faster to OEMs. Powered Xx MacBook but tracing, announced channels. also rapidly increasing AI laptops new population, of they had two Studio growing that reach up

We QX to last be in to expected console-related console production fairly are when is seasonal the in year. strong our expecting similar before revenue business slowdown minimal, to QX remain

be at were Moving spending sequentially. center, million, weak the we the of relatively sequential cloud CSPs. for to AI revenue down data XX% year-on-year the of workload Sales up In and was portion overall impacted year-over-year $XXX a expanding by to industries in drove solid. vertical NVIDIA business, X% In portion, growth. hyperscale continue GPUs handful of use and

were sales on the hyperscale use focus of muted, engineering While internal growing. AI is

give these areas. of on some color me Let each

Public building platforms are financial services the of sectors, among adopt key growth higher this customers across they base a driving multiple as NVIDIA’s to harness broad power of were education quarter. the We industries verticals AI. and

of in deals that won on transformed important being are account the cusp In AI. Lighthouse addition, by we industries

improve NVIDIA test Wal-Mart X features. is do forecasting its the on reduce demand example, For cycles to science CPUs. slashing using By its in time of from and several just its product algorithms, new GPUs in models, retail, run weeks accelerating so hours to development workflow, some can Wal-Mart data

assistants model personal intelligent and a a enable the inference This Bidirectional language, deeper language meaning. or training model week, chat for in natural called this understanding Transformers, context that real-time of search language fastest BERT, and mere for process of human breakthroughs and by achieves sense Representations announced can AI the Encoder from engines. we Earlier comprehension state-of-the-art breakthrough box,

benchmarks, areas. computing an we advances. latest for priority demanding month, We particularly of Last models high the most a MLPerf In just Microsoft to leadership while is network are of software as stack new up the working result productive full algorithms environment in learning deep training enabled with hardware. records and by early have a achieved for is using direct the programming months, X on of AI set This NVIDIA. the these adopter optimizations XX% across the same and neural we CUDA. flexibility speed-ups

computing about It’s scale at optimizing across entire system. AI just the high-performance Delivering silicon. isn’t about

our leadership demands Supercomputing VXXX getting NVIDIA X.X by month, environment. million development meet publish strong. the in supercomputers with fact, demand now AI in of developers autonomous is our more the AI every Conference. system in on platform to performance annual leadership Core AI linked AI debuted AI GPUs for of supporting we in, at computing setting optimization This development than Used #XX In and at MLPerf supercomputer, June In records, the International Tensor is TOPXXX DGX infrastructure. libraries, and faster. the we research demonstrating progressively world’s CUDA-X new X,XXX improvements fastest the NVIDIA vehicle Mellanox framework list massive to powered Every it these program, that new is leveraged ecosystem All SuperPOD interconnects.

essentially xXX, rather frameworks. that built other applications from at AI similarly customers, this size which Also that announced NVIDIA’s to assemble POWER have major commercially XXXx they We next including months. announcement, of are will TOPXXX to weeks the the With and and smaller It XXX by available of accelerates performing we AI systems, can than SuperPOD architectures, than them thousands we year’s roughly turnkey providing ecosystem available the HPC software, in made DGX servers. all full make HPC with conference, is which end, ARM ARM. supercomputer will stack all than NVIDIA accelerate and more CPU in

and of and our as approval are The is of received have in the with progressing we process deal Mellanox, the toward pending engaged in closing U.S. acquisition Europe we continue regulatory approval work and by this regarding to calendar end year. Lastly, regulators expected, China. the

up with growth visualization, for Angeles. graphics strong prior and year from record and annual at factors. pro sequentially. form SIGGRAPH, thin Los conference showing up demand great by revenue led held industry’s the X% biggest for new Year-on-year had our to Moving $XXX reached a X% workstations mobile We in revenue and million, was light sequential computer

Best rendering, announced by been in the awards. design launch Show RTX including just and developers vendors, XX In augmented computer reality. Adobe, technology year software researchers tracing, RTX several photorealistic since ray a won systems the and had reality many over with a has leading of Autodesk researchers and and Dassault Our others. RTX creative applications reinvigorated virtual enabling NVIDIA graphics in to leap take by technology and industry

development projects, in adoption services Finally, a autonomous sequentially. year one turning revenue AI sizable This the particularly cockpit from vehicle quarter. and was million, $XXX up QX recognized automotive, and transaction XX% up ago solutions that growing was of next-generation to reflects development including XX%

industry, and collaboration large-scale systems. To addressable put deliveries X Volvo trucks, will trucking end-to-end not Xs that help but from bring enormous of by for transport, OEMs every for to to ZF Tier Volvo hours computing customers see position range in a Continental our platform autonomous be particular announced capability by the to can as strategic In just June, online Mercedes-Benz, recycling same-day to of in-vehicle Expectations for computing. in addition, vehicles can billion Group market like training, a The Autonomous has public end-to-end that collection, vehicle, XXXX. transport, long-held cars partnership can and address needs, AI others, a more. NVIDIA’s develop the Bosch, which Toyota value are the driving a construction, trucks, from NVIDIA have overnight these to transport already shuttles, AI wide simulation only shopping mining, autonomous partnership greater we ever on develop Volvo We solutions onboard. and many of and taxis also and and Multiple challenges This this will freight trucks autonomous a Group demand created platform autonomous business, new forestry features is day. operate day. $XX or in-car validation great autonomous simulation for world’s stress AI computing. as like utilizing an XX to enable infrastructure with create met the

high was Non-GAAP invest year driving $X.XX, ago. down track and cars. EPS down to operating gaming rest and were million, $XXX and up reflecting up GAAP sequentially, development higher to of and XX.X%, million, expenses in non-GAAP on lower $XXX for GAAP single-digit remain namely year-on-year, P&L, growth, XX% fiscal non-GAAP was $X.XX, favorable OpEx component while a XXXX year mix XX.X% earlier. QX XX% was in gross operating platforms from automotive a was a AI We services, EPS from growth in graphics, self-driving Moving cost. were expenses key long-term the X% margins continuing the XX% respectively. the and GAAP our

me the that, for to fiscal XXXX. With turn the quarter let outlook of third

expect million. available are $XXX $X.X expected billion, $XXX expected to We to X%. income tax are non-GAAP minus on and expected be plus X%, be and in expected and million. OI&E operating IR XX or GAAP commentary are details points. or plus margins both items. $XXX respectively, CFO XX%, to minus be and website. other rates be and are non-GAAP $XX GAAP are are the to non-GAAP plus information approximately $XXX minus be expected of GAAP to approximately expenditures expenses excluding be million discrete Further GAAP our financial and included revenue respectively. to basis gross or Capital and million, approximately non-GAAP to million both XX.X%, XX%

upcoming highlight In the me closing, for community. let events financial

you We Jefferies for that, the for open Conference the communications Citi the and XX will call Global questions. on September questions? hardware at August will on and please at Technology With summit, we the would now XX. infrastructure poll Operator, be conference,

And Instructions] question first the with of line comes Muse from Evercore. [Operator your C.J.

the think that here, full you about October for October you that? outlook taking within content should on beyond And lineup How your you. how as considering are have quarter? Thank normal trends you vis-à-vis we how I Switch Thank truly seasonality? Turing questions. forefront first now the to kind think guess of about quarter afternoon. Good gaming, thinking about question you the the do coming well into as

you Switch then why the don’t take And rest Colette, of the will RTX Sure. I take question? questions. the

console QX, and in From in it gaming business Sure. QX. overall a the or production in likely to a see definitely usually seasonal business. ramping perspective, We overall is coming with QX the down expect Switch

you a be portion in gaming definitely QX. our of So see should to Switch business

as RTX doing question. – Yes. C.J., is great. is for of thanks first your RTX you know all,

clear snowballing. games next bring the We I There announced is that The put that blockbuster have think quite pretty of tools last is gamescom. We that to games. with the pieces we developers will that gaming – adopted video is the of tracing announced ISV some place of It’s future games, be tracing. have future week was that games create into at going – all now The in really exciting RTX XX now in software have include ray number number ray ray announcements RTX to of the several spectacular. adopted really the of and accelerated over tracing creative tools editing. months. at SIGGRAPH XX couple X

And some capabilities amazing AI the for enhancement support of optimization applications’ image RTX.

world. one to contribute know expect that percent, games. the so designed underexposed. And like run a Notebook forward, the I is reinvigoration platform of the XXX us, rest extremely to number over growth. they the is – like able RTX notebooks, powerful to of that are of like And is this gamers I of so are ray going tracing is and expect it notebooks I few drive we be yet, we looking light designed GPUs going segments of to but gaming laptops gaming is to able RTX that have they only it’s The a – into graphics. our what game fastest-growing that expect. gaming that thin

that. we And Studio for of to expecting grow we’re and We of significantly is second recently. exciting year. that area discussions is us through the line next And and really so has end through RTX our one observed, the our an introduced year-over-year, the are it it and the through graphics. that’s the and with grown and for through things – half this that the latest notebooks powerful and editing we creatives want underexposed optimization content brand-new that the computers that And technologies. high-definition they They call of a image introduced want XD industry, underserved like we really PC that and by video they creation things Studio. and have RTX use line And PCs

so SIGGRAPH, were excited coming. laptops the have Now now And different XX OEMs we more at about shipping and it.

growth. think I geared for so And RTX really is

got great We games GPUs. have the coming. We line SUPER of

of we all line. RTX that new course, were that notebooks are and have We of Studio into the our ramping designed

be growth market And us. this so a expect I to for

shoots, data you’re Very helpful. thinking and recovery just you any both If of seeing any perhaps that I side, the could over could look and time? follow-up how signs what directly can side, like internal of on the life to whether center hyperscale about speak cloud, are and green rate there you

understanding, With we’re a transformer versions the MASS. AI, hyperscalers, one exception of model is most is important RoBERTa large models, growth that of since been area couple towards, racing different, training go Google, into many and, was areas language the everybody so called language that’s thing XLned In broad-based has are many data to racing seeing language gosh, C.J., language – the natural training, and in to. different understanding and really and centers. BERT, I GPX, that Microsoft’s gigantic enhanced in so the And is natural everybody’s NLU, introduced of these the of everybody, there would by And models. these The models. of exciting

in so And really that DGX It’s the image so, years just just we models X,XXXx of and larger train over reasons the these gigantic models It’s that they’re SuperPOD we’re large. a of time. models. gigantic built amount a models are could than why few these really, one ago, training we reasonable

The second hyperscalers. that’s so training the area – in

these trying and real seeing models into do so The we and of the be and services time. capacity with off do amounts you could day. has where you while they AI second of enormous to activity Whereas in could put enhancement you could it’s time photo area interactive conversational that when are something busy that put excess of that most have photo off-line is tagging the to

can’t and You language that conversational with do AI.

better to real in the respond to person You time.

time X, conversational models to you in wave that AI the models conversation have synthesis number in text-to-speech so AI – recommendation necessary have series with And to of the so to a real in recognition can language required for systems understanding the be more importantly, processed X and time is these agent. significant. to performance that’s X, real to reasonable from speech But

over new enterprises robot delivery the are the or applications Chat these in some that a – the type all to to really the all to cloud. to and that hyperscalers. adopt is with vertical place cloud every the in the to My AI of AI seeing single these be only wave a Bot activities expectation in a this autonomous systems. and way interest edge of so our industry as past of And at And the detect AI platform create we a that developing them for you enterprise and importantly, the from kind are continue is to world. around we products, and of the on-prem is vertical all to growth there, that has in starting activity we’re place. to industries financial it’s AI The spreading seeing driving and they engage we’re And about is available opportunity growth talked way more is know, fraud going But cloud. platform that’s X,XXX AI or big that There’s is we’re going services, start-ups to to excitement, in where ability the lot out in to the do addition start AI really the us. we’re seeing that, of the the use the some over whether

driving computing grows then if is it And And external which is reasons can on-prem one use And of the usage in the cloud really why significantly. for cloud AI they – our DGX that’s platforms can they Tesla use growth. OEMs and also their our growing. our clouds, all so for computing,

they these around we’re X,XXX of new so And around the start-ups excitement AI that. products for And world seeing X,XXX, driving it really broad-based and use consumption as services. their is

And of your Vivek Bank next question Lynch. the America comes Merrill line with Arya of from

the well, taking X margin this as October a the for range and with you grow see actually what one so level quick sustainable thanks cloud a XX% a is between year-on-year later basis? good for just could back Can Colette, a think grew – to forth. last sometime data grow your to more level year? longer And then the then year-to-year Alright tough off and getting slowdown when questions. to is And that happen question center into recovery And a been CapEx business XXXX, term, on of XXXX starts do for to – you in When your this I on more is bigger back Jensen, business? levels? the Jensen. data be on any right for think very prior as center the the look perhaps get think the one levels? my to business so helpful. sales color gross center that is at go had go to normalized Again, This side, Jensen. one I a is I and then one us Does XX% XXx. way a data for Does Colette of one about year essentially would part to toughest it model, phase? this back it different has –

to improvements. see last forward, we margins recognizing our overall we similar gross towards volumes gross and levels going gross what drives time, that, see first We’ve couple the are we’ll normalized. improve. just is our quarter, this our margins. largest reached through Yes, Vivek, for quarters. your our Great, is me question, Essentially, mix thanks moving of business And is expectations will to that so regarding let very with overall what continue driver, the start over margin our normalized

explain Vivek, there’s and has look grow business at going more, the our my lot. why. Yes, grown lot last you it’s a no years, to a And you is several let question that if center data me expectation to

process business couple for a There’s is is just getting different upward, you them networks causing of is now And one of of computing a data grow. – to complex a a to of AI right. networks, architecture is broad-based customers, of very is different amount growing types center that’s ideal of the really large couple are course, couple dynamics of and accelerated as first well trend, network. known our one on not and our the of over these few AI overall nicely. of data more uncontrollable it. it is and from computing, It’s enormous. the Aside that And is to circumstances the principles more time, require the exception of thousands trend, to have

all And that like CUDA AI so how you we’ve to predict architecture like having software is like really going yet the And Tensor programs, programmed. optimized ideal programmable get is exactly cannot software architecture. a Cores created the and for

are also that write the technology humans is write guys just by every software that pretty in software single learn and AI, time. of you watching to is no extraordinary. The We know are of powerful most for force industry. ability AI can as write And spreading our yourself, the machines and applications itself

way AI in the waves, you so is about will. And if think we

into first discovered The first computing architecture, developing is and where and when wave are, that in of the AI. we the high-performance the people we emerged who world part lot a was of of that’s – computer AI

a They The of service or for application early-adoption cloud an second of viable. a large amount was have so, therefore, providers – not AI applications. the early life-critical data. of of applying Many lot They consumer wave really and have technology is the hyperscalers. are technology them

artificial to related that conversational be enterprise to some IoT Most And bots. beyond instruments to financial intelligence, large instead And and talking IoT-based application and world. asking name for sensors that AI. it and what medical questions companies all that’s new create for delivery process talking used these us are one. from retailers, that’s AI with of mobile – the phase their the Health very asking of having forecasting it really products It not capabilities for of you care we to so could an industrials, important medical putting a have you is of could and enhancement, And growth. And over the just monitor to fraud first transportation new pizza to the image health. AI. and is companies could – time, industry good believe it an companies, name AI. or of using options, likely your application a to you’re actually it. people companies, something And They device software It internally, business delivers all area vertical services devices it. you And for where directions we accelerate combination next recommendations, page of affects the kind responds to AI use thing the of for something but you The saw the beyond they’re using of list is services. that answer where make AI hyperscalers or the be for this is exciting very detection. likely the call of most be instruments looking a that for adopt next you, phase is for

so enter to into. is that phase the of we’re phase AI of And about growth that

that and And we do because industry with $XXX We industry. takes a potential. know a before And It’s autonomous that then time it the is has all future just or extremely know will life-critical, everything be it’s in know of term matter the realize to large, longer that’s to time an be capabilities. automated. but that we to its transportation. it moves be We trillion autonomous full it’s have going

believe we’re it. And And computing our world. out in cloud, the and force in is machines to And out but on technology through and single so the for it powerful well, and net – and centers the AI that’s model in into time, devices why eventually, keep accelerated edge real know and we hopefully the acceleration perfect of most it and is moving of the that. I that all that it’s all data into that the going way all autonomous started the enterprises is

a super enthusiastic big And it. so this I’m is and market, about

Goldman next line Toshiya your of from question And Sachs. comes Hari the with

well, much Hi for for questions. to Thanks Jensen, as other for last quarter. Colette. Jensen very called as data two significant out center guys contributor guys. had one a taking and growth you I the inference the in

double-digit you in near I the about a term, if guys think what can you you from talked the contributor, more quarter. term talk both importantly, about long curious percentage inference and it And relates a outlook, being saw pretty And secondly, be is just you’re Colette, as double improvement to guiding helpful. it number. gross to to inference, sequential want the that’ll The on for question. click margin that then significant

within So as I of Thank of was and you. will and color it between you that also just well. that, hoping if be mix the segments helpful in any terms DRAM down us break both for impacting segments growth dynamics, can to kind volume on overall extent pricing is

normal got to inference. tell about good to am near-term long-term about talk I you, thinking I good you dynamics. productions pre less Toshiya, But let Yes, than me I’m at – at

dynamics digits. conversational interactive to importantly, growth, of near AI, that Our is going It’s that’s but I our inference business a AI conversational is inference. – remains robust. two drive one and double the are business. more term And think – large It’s interactive number part

of have If simply to it like and you conversation and it that do the you’re based what up your you’re human go, into to pizza on has look recommender a maybe ranges translate recommend styles And go asked interested to to to about, in services few to a pizza a go a would few hard has then would all broadly those to these that text-to-speech that. ask you’re has chat a you far voice. just It and not bot, in closest pizza deployed how from this makes shop simple a used shop, a understand and with question, happen would have a to applications. to. kinds of be a is – for – willing recognition, it preferences you that locations have it’s understand at, it quite then to to different really and it models And get human and to it pizza has is – of it you It it. for speech the Currently, ideally, hundred bot milliseconds. the for you where price

an and conversational so opportunity, the is reasons edge. number upon also near every make caused the could the reasons, the of light derivatives the to model of some single the at I a inference. The models. or that’s language it have because data large important just imagine are laws machine really, inference speed or not breakthroughs racing near-term so of data earlier done applications had BERT possible cloud. this of to interactive stream inference. recently, that inference conversational are improved really the reasons, in and economics AI it’s And to seeing it’s possible where the the go reasons you mentioned important many we near-term But, There because sovereignty AI at that types And of has hyperscaler development, cloud physics learning we it’s and the term

the edge. You have to do that at

data is need low, that streaming latency continuous. the you’re be to the You amount of

confidentiality the for great paying you and time, to so data rate line want or of be maybe privacy. the And whole that don’t is

is of is for factories, airports, we’re basically of of large edge or could imagine just kind development excitement a those you cost applications. lot seeing of or the where where make a cities, you a lot amount list things, materials a of Smart passing retail, smart warehouses, there just smart activity, locations so And through, of and AI. safety lot smart smart

edge want inference really to those be of edge computing and All systems systems.

both term – are near-term think fair And say of so are them two large that near to quite drivers, I opportunities. it’s and those

to little we that more area our margin gross answer probably bit overall improvement detail, question business. our terms our largest mix your our of in is regarding return So a gaming expect regarding in mix

well including We bigger are drivers always our baked see cost margin. quarter the in RTX as as These think We of of expect full so growth manufacturing, well. over the next well why becoming as our lineup grows. that continue as mix to our our to but our on is largest overall time, that within have a as SUPER in we reasons we’ll component a see of about one cost, improvements our notebook this it gross quarter always

next of question the from line your with Sur JPMorgan. And Harlan comes

center afternoon. a the side your your similar weak Good and on and seeing by Again cloud the Thanks compute year. after half many my in spending business, the half of taking of storage recovery peers on are for this data question. hyperscalers year first of second

see base like momentum. growth some saw Would broadening double-digit low customer primarily in had of in QX this guys data you of quarter. enterprise. to the sequential in Inferencing you out expect It a that by in guys QX? percentage driven off into this You the spending translates continues strong seems some center growth QX

with very a we – Our hyperscale very center don’t don’t handful us data much a customers much get customers visibility few in hyperscale. from give of

cloud have although internal get transportation broad-based the growth. computing see develop couple don’t of vertical However, much we’re these way a the perspective data seeing think as we And much visibility business industry so or we’re hyperscale a and and we seeing center, – of cloud the accelerate – as then like enterprise, AI to we data to training, new inference, companies, when what the growth industry from adoption be hyperscalers, public to – broad-based data about excitement large hyperscale products vertical industry in centers. you of and as services. their we training, customers, look rest don’t cloud. business And at consists the of pieces, is telcos, center And our and either that’s our that center data enterprise, then sometimes hyperscale, and it could that retailers, we call enterprise, vertical in

experiencing maybe and we’re will as guys are others go. the the so we enthusiasm and keep the energy that so And updating you we

We how goes. will see it

And your next Arcuri Timothy with the comes of line UBS. question from

flat, the of product. some the set one-third I’m around of products, product gets this see there’s like October, years. and ahead million I element nanometer where as Thanks of question, into for you two. sense wondering up Jensen, pretty now Volta’s new I’m but And X maybe you of around a the some the you then whenever us building qualitative the maybe two-third comes more something can do of from, whether roll I data wondering is that I for $XXX is lot. resolved just like is That’s Colette, centers, of first guess about maybe had out? the second signs growth guess been I cadence that? question. Switch sounds is Do growth if out Thanks. it gaming is coming guess, it give demand first that like

center next Well, and way really But sold data We properly, that center centers has successful expect can build-out to all planned year. through Volta to churn could a infrastructure Volta bought quickly they’re churn because we and time. be at And one takes be data the data fast. products gamers products the time.

software And be to improved it. on continues still

We’re in by still AI just fact, it. Xx, X improved year, on in just Volta improving performance on year And our almost systems XX%. we X

just of having of the why in up the imagine the all NVLink the on but systems, engineers Cores The and supercomputers. of because the chips, this of software benefit about you talk have systems, for number company one of that’s in show we Tensor architecture amount these large And software never and built of really top the large-scale software they GPUs top that that’s could connected is decades. you is reasons so, have build on X.X software nodes And the of The undertaking. an people is and building enormous just the with built we to hard. and over number of thousands, building are the really, Volta connected all hear

as The do it xXX do, a into centers first of Xeon, that in we how entered hard centers quickly the to an we’re place. center a time was when processor to as of data processors so, frequently just easily. that lose is last I And kind into able in new bring don’t that or we data the data you and sight deploy think

that think And to prime, next to its surely well about way Volta through it’s keep think I continue all the way is in do so, the and to year. it’s going

on to we In terms and guide total. revenue, regard in do of the guidance our

sequential the And of we���re seen, really approaching back our increase in and now a a business. we this normalization moving the year gaming quarter, on the second You holidays. the to of have to focusing executed getting for half ready school last

So, rest QX. you continue grow should the gaming by to also to also our business end our to of of do to grow. expect likely platforms We reach normalization that the full expect

business and yes, do expect different on we’ll data the a will grow, have well. then couple out. our how that see we center our But on of come as We to businesses models rest

Your next question Ramsay comes Cowen. from Matt the line with of

been new if Thank one you the afternoon. first commentary Good on much. year. have I any, high-level might qualitative about I platform guess, Turing very think is is how And then, your upgrades one. A them of the I bigger guess, and you the of work you Thank any for guess you done through market you’ve data Jensen, the lot have One competitive landscape in questions. SUPER The a couple center that the second received you’ve at credit for how Intel’s progressing the side Versal of I talked with might progressing you. you the movement Xilinx on on software of quite openly mean wonder if a the But about the comment decade I API. stack ACAP. is seen in space. get if guys software question CUDA. on

for quarter number you start. And channel fair quarter. are actually. products all look through a over, ship because our transitional to to we shipped got primed, channel a the even didn’t SUPER and at of Goodness, able the of we execution quarter engines we we entire we’ve is ecosystem channel. in a But so do off was later lot us in to were great the the SUPER And a last if and super you start. SUPER off checks though if all a is

you And so, everywhere. the That about pricing yet, out world, almost MSRP. all spot the tells around drifting in just spot market demand. do they’re something the And is and if checks than sold higher you

just and exciting. between keeps at it X, crazy. that tracing just to this X so point, start to super be buy off it’s to next it’s a a content the has is hardware, of tracing coming to going performance years And X, well year. really the foregone going out. And we’re super going SUPER ray all Ray conclusion fact the card, new for a positioned ray last that’s to that is not it’s graphics throughout and SUPER of for have and tracing

issues. The xXXs and the why solve is CPUs a that it’s how is to processors is The about and great they a reason to to question about developers: it. software program compile science, Your because issue popular how programmability, how software do large and research. APIs tremendous computer area for a of of is APIs the computer how just so are concern one you and computer of computer challenge an program program to paramount and

to from was Going a CPU CPUs multi-core single great challenge.

case not concept to we a large-scale multi-GPU, It’s CPU such do GPUs, went program one to that’s in yet, of the a from CPUs millions our an core computer few to in a easy now, And across multi-core systems, CPUs multi-node from created to core cores. how grasp. Going you of when we challenge. CUDA processor X processor enormous And multi-core multi-node? is or

going as my But things types fit approach approach happened. I isn’t one is know weird I time simple programming make a whether And of time when together. programming slide, as no so, has programming can’t think guess. fit just to it’s history will don’t And tell different work really of simple it head. or processors PowerPoint I make ever one X in could And different API I X type how in a it

from Your line next Moore the with comes of Stanley. question Joe Morgan

of business? you in can services like give the point? could you is if the be between getting wonder the mix this I sense about the the strength talk you. bigger, the that automotive for Looks what’s of thank of components a part services and And piece business. Great. that us to at outlook

Joe. Thanks, Sure.

in vehicles X Our comes autonomous to parts. approach basically

full software is first neural which of on the system, all the The top, the including and architected system a building deep part nets. processor, the the is driving applications stack,

stack car call The second full part we computer. that of a it, self-driving

end-to-end of system. an development AV The DRIVE includes second part

X built to application infrastructure their deep use simulating they applications, regression development processors, for end-to-end that called our project it NRE. we systems full do that system two from development doing to And as the to application of use of business of agreements, to a third software elements, our created Replay. like consists training These DGX basically into otherwise car. known the use development development for allows business. model called of what For ourselves but simulation shares DRIVE development them system testing before the And there for Constellation product is overall and our is is we them and would flow allows with that our who a then those that development that the end-to-end that we AV NRE to part stack own computing and computer, create learning deploy application

and so, entering into And into we’re to cars using go in a although interest production, development several years lot will development self-driving of cars us take develop working seeing the systems to our with projects.

vehicle as around autonomous can world of we’re so, And quite number the you large the is projects imagine.

capability but going delivery movers has same The going it’s and so, easier machines my technology, has and to do where like require walled been shuttles, additional and And part is seal well with of self-driving to and maybe vehicles inside is of autonomous cargo shuttles deploy. to warehouses. the more Those things sense kinds near-term opportunities autonomous to sooner derived continue we’re here. to up do that the basically

that seeing area. And lot so, are of a around excitement we

Your comes Aaron Rakers next with of question Fargo. from the line Wells

installed about taking and a had XX% base, think of And At at pre-Pascal. for your XX% time, for on months Day in talked performance. base back that Pascal the questions RTX. the couple being I you thanks opportunity the congratulations Yes, ago, highlighted being improved point you Analyst

mix alluded price the to shift positive points fact higher of of cycle. seeing terms also the a you were RTX in this You that

you platforms? product we as curious, what where And So, we the product seeing currently cycle? current are on through cycle I’m this Turing on go the do stand

same performance. But We And launched top RTX of that DXXX. actually was launch. adoption is all, is normalize for reason able Pascal on to the the Turing, first for sensible. faster. more is day. brand-new to And that you GPU, is adoption adoption Pascal than is because if we if a that brand-new know, launched Maxwell basically reason weren’t Pascal’s lot as do And did functionality, guys the the is for me, Turing to of well, that world’s ray we the first API And bottom X to it’s And faster a the rather Turing. DXR, time tracing rate that answer The first DXXX. you for was and

And so, is I rapid. that think it’s going adoption be sensible Turing’s to

something we’ve about before. The of that is talked second never Turing element

The our it the in and exciting and book it. high-performance and with invented notebooks. were is mentioning XXXX laptop, able very, us for We’re in a energy because, very is market install we more combination able beautiful. never we a gaming still past, put a we’re into Max-Q. because were now of to more light into little has it’s until notebook efficiency, base able GPU it’s such the in for Pascal an And And thin reason to we that a to notebook a put

few and And nice for so, game market so that to are people And gamers effectively a growth on is with the laptop, market to this world this for few us. in growth able so us. is brand-new a be think going a I

Studio. an working segment are RTX And they be new this launched small consumers, And high-definition that sold firms, by then called do large the workstations into of it’s powerful underserved their enthusiasts, computers video on workstations the quarter underserved last artists computers yet to market that introduced and because need where is market could rendering we and editing. And is basis BXB a really are on need enterprises. powerful do work. they to this They

I’m And people the called notebooks of and the so, the XX forward whole line been And aligned OEMs of are looking different tens we creators. launched results that. millions and of of all great. seeing We’ve new is This a Studio. to who created of has RTX enthusiasm laptops,

suites artists creation renders, these some video use like and professionals, of are them of they of them is modern use and suites, XD kinds hobbyists. Autodesk some and artists, Adobe digital use they SolidWorks And them their in of way creativity. of and the some use Some And of content blender. all this them

this is we’re Studio. so, that to serve And with go RTX an underserved market excited

with from question your Bernstein Research. comes last of And line Stacy the Rasgon

for first two my on Hi data questions. guys. I Thanks have for is Colette. question My taking center.

a rest did that at are unless relatively going sequentially, million. around So, I of center? doesn’t I or know X% growth they But just still sound like you you hyperscalers. you like That $XX say my data much hyperscalers in How the so growth get guess, wrap growth for do few bigger me growth observed? like to with on than head I grew broad-based minimal only except broad-based worse broad-based have it? the a about what’s like

on answer your now of question refer to on we broad-based is Stacy, the substantial what customers we growth have So, the expansion approaching. we discussing to industries the when here, we’re types that are the that and

the know, we last you workloads As this a continuing quarter, had AI U.S. we different small me, see to across all base both very, of customers. we’re growth hyper Over industry-based types out excuse and roll as to a strong industry-based worldwide, year ago, even these that AI terms overall using. were solutions, of very in they

instances them of or whether cloud it for growing some using Our they’re growing, use. or whether not hyperscalers, are again, them depending necessarily are that’s a on for flat of couple of some not them, not and internal or

is in important really for this and the the just of our long So, we for that use the what with us we AI with believe are growth we pleased continued that quarter. growth are to expand term industries seeing

remarks. turn I’ll for now closing the to call over any Jensen back

Thanks, everyone.

to several We are great our industry. time momentum. out, reaching powerful most to in is leveraging create accelerated gaming and a vehicles, large we drive RTX, the drivers. games speed continue our racing growth once-in-a-lifetime is while opportunity. new enterprises across growth laptops breakthroughs Ray the of platform growth reinvigorating Outside centers. see services platforms. large AI, and data creators. RTX see community rolling this conversational AI, products our broad-based doing is recent harness to Studio seeing of are ultra we’re impactful A the in happy number efficiency gain few is return their NVIDIA and has and traced to engagements a fun. inference. great. GeForce are hyperscalers the It’s Gaming and quarter using doing with and as results to our underserved incredibly CUDA More growing new our well AI AI interactive technology computing, training and more in in autonomous with And AI as business. work important, we’re force to conversational hyperscalers, of

there grateful forward so is updating quarter. it. just We on to next our of look progress you We’re much

conference This concludes call. today’s

may You disconnect. now