NVDA - Earnings call transcript Q4 2019

Due to licensing restrictions, you must log in to view earnings call transcripts.

Operator

Good afternoon. My name is Kristina, and I'll be your conference operator today. Welcome to NVIDIA's financial results conference call [Operator Instructions]. Thank you. I'll now turn the call over to Simona Jankowski, Vice President of Investor Relations to begin your conference.

Simona Jankowski

you. Thank everyone. afternoon, Good NVIDIA's for to of fourth fiscal XXXX. conference welcome the call quarter And the from Kress, call Huang, on me With today Chief and Financial Officer President President NVIDIA Executive and Vice Colette Executive Chief Officer. and Jen-Hsun are our you on Relations to site. being Investor is that like webcast call I'd Web remind live NVIDIA's will webcast our replay XXXX. results the for for financial of conference fiscal the The until available first be discuss call quarter to content The today's NVIDIA's call property. of is written reproduced consent. be It or our prior transcribed can't without

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

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

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

financial find to Colette. on commentary, measures You these turn our over call With to the Web GAAP of a that, in can site. which financial measures reconciliation our to posted non-GAAP CFO is I'd like

Colette Kress

Simona. Thanks,

was weaker results QX gaming down are XX% Full year-on-year XX% year and $XX.XX Three XX% the down of million revenue heading with by quarter. up the $XXX sequentially, year. know gaming you expectations Revenue billion, revenue earlier, XX% to from revenue billion. in fourth business. factors to our the $X.XX year Starting contributed line into our for January Full lowered in up we decline down previous XX% with than gaming. was primarily decline. As the pre-announcement. and year-on-year driven quarter a billion, was was QX XXth our from XX% $X.XX year revenue guidance

post to inventory of in the sell GPUs in us order excess channel allow inventory crypto to caused shipments First, channel to reduce through.

in one timeline inventory We call. our normalize on we expect previous to in two channel to earnings line had quarter QX outlined with

may China, but third, than performance deteriorating customers of ray-tracing deliver quarters demonstrations price after for particularly and demand gains. the and its certain technology underlying their we architecture, in for waiting high while The and economic in distorted actual sales conditions, GPUs; revolutionary obscured using the a our Turing These has and to over trends delayed consumer end its and RTX innovation products our expected impacted Crypto the XXXX challenging including time gaming volatility and RTX been last mining new architecture. lower few leap purchase GeForce business demand line. Second, model. lower were GPUs macro some AI, our real have in have the launch for points of of in new further XXXX quarter-to-quarter significant a with gaming effects and the or the business trend

believe we what long like. into visibility the some you business give to try live me Let looks

include across desktop rate run gaming is crypto the of last to year, in you revenue our of a business get business period it our know, the at gaming, that To sense demand. and look As products. gaming console underlying gaming desktop consists gaming gaming notebook helpful doesn't

of in QX. normalized the the from a QX year gaming so allow last and inventory, million revenue same look demand represents quarters again per QX channel $XXX down to will the a normalized gaming that average, gaming and these at end be billion. QX consoles to revenue this [indiscernible] our Therefore, gaming desktop of is quarter levels. four gaming over desktop QX higher have where approximately in turned On quarters demand such to amount channel hindsight, level work is averaged are exiting In and quarters, to total this products of with expect current desktop excess relative close of benefit this last inventories lower QX end out Notebook year starting We four across gaming the relative believe believe of shipped shipped we do four Let's the normal rate period. for run at to of channel or these year. about we we To we business to in quarter $X.X QX, amount $XXX to be. million.

As the behind month, channel delivers and business business to we a features [$XXX] at XXXX a XXXX point to we gaming rave received great reviews range The we tracing into our of and CES The bringing full the off GTX of our XX% to expect tough AI mid the business. be over while announced performance a have launched the We On has bottomed. for gaming gaming. our mass recovery given real market time. XXXX price [ph]. At first the RTX ray slightly time compares QX, growth basis, XXXX and we look improvement inventory start. to also be is to correction GeForce Turing's the the last down half expect with first Turing notebook in past us and the year

a of the XXth.This XX GeForce market of to the which announced Featuring the the laptops, Turing notebooks powered Turing new PCs. than of gaming efficiency architecture record the the plus more gaming we are is addition, number last year. performance based desktop able on in In deliver [indiscernible] available double energy became of laptops January

of be We're GeForce see ecosystem. fastest momentum gaming. to pleased to growing segment laptops in the continue RTX to growing also the We expect

content developers game amazing creating architectures' are the Turing As take advantage more to of capabilities.

is and week, Battlefield games, soon. DLSS Just coming and Metro X this Exodus, available blockbuster technology two Anthem in becoming

RTX technology. Justice but of restorization can frame more gaming. DLSS, also Pairing integrate to CES, to of at and visual, DLSS demos tracing the next worlds. showed both comparable featuring rates beautiful the tracing This is with games addition Heart much of ray generation ray pipelines traditional Atomic plan cinematic In the best large and provide and

in our was a solid to gaming channel us inventory the of on this the challenging correction While look behind our we putting forward foundation Turing and business building quarter architecture.

$XXX was geographies. center. XX% and to XX% Full Moving year XX%. Revenue year-on-year was up The and data data down million, was and up decline revenue verticals sequentially. strong a billion, across QX $X.XX sales markets broad based center

cautious As due world and rising not the the of economic quarter in became customers close deals around increasingly progressed, number to January. uncertainty did the

year. The accelerated strength both pause the at remains computing believe intact. temporary. of paused several In as sequentially year-on-year addition, the is platform cloud hyperscale NVIDIA's purchases customers and declined of and the end We

can in continue continue for of and We expand to And programability, deep computing lead performance platform. the our CUDA's value the with to scientific industry learning. we

data called cloud workloads, in example, NVIDIA variety objective Europe, recently December, available Tokyo. TX computing, TX inference, in benchmark and in deep beta For third-party models India, In training in. high-performance AI six including the its learning. And the Brazil, broader Tensor a centers the categories and analytics a announced of CUDA competed became data The GPU and that announced Google the Core NVIDIA Singapore graphics. RAPIDS, acceleration learning universal we we January, And test for are is U.S., our GPUs first in machine learning now top accelerates learning, captured deep machine training that available. and stack analytics that MLPerf for Cloud data spots in

the remains environment, segment year. data the the until visibility spending forecast low in center don't Our cautious meaningful later in current a we and recovery in

their with closely working center of business, are addressable GPUs currently market into world going integrate hyperscales and drives we However, software NVIDIA Inference significant our to opportunity around while a forward. XX% less our production than inference flow. TensorRT Tensor represents of Core the expansion data

portfolio we in financial create and we have Cisco, machine capabilities industries our product also to address through stack, that for data services. such have Pure retail, have With services, RAPIDS systems it consumer technology as done customers and their by to Internet to analytics data NetApp and enormous and channels. easier We deep IBM, be healthcare vertical strengthened our as automotive, NVIDIA pre-integrated sold can our software and can IT requirement, learning. accelerate analytics an with Storage And global made and go-to-market that partnering learning adopt

visualization, XX% year-on-year. up million, and up year $XXX X% prior from $X.XX reached was to down revenue the billion, XX% revenue Moving sequentially. pro Full year

designers quarter, New workstations work data enabling mobile of This HP, VR, to workstation Lenovo, as creative science, significant well Quadro XX including applications super remain key Google, media improved announced like enable for the to with the first high is applications AI in and Boeing, the business. light millions and robotics. LinkedIn for of with will photorealistic including the AI RTX Dell, as our the them creative partners key week most we need workstation. dramatically and years. in time past RTX availability rendering, XD and productivity. with GPU upgrade wins and artists thin drivers for We It Quadro had the be resolution Toyota growth interactively and for

plus world's largely CES, million, and and outside Bosch, by multiple system XX% on-display AutoPilot, To The deals, driving increasingly from features and capabilities, were camera generation decline processing we offerings, increasing deliver a Veoneer, legacy Finally, and turning functionality up performance, surpassed revenue car seasonal. our our first to infotainment. perception up X% the solutions in declines car X uses today's road capability. ago adoption DRIVE booths. was offset next at development down sophisticated and of XX%. revenue the in million, was deep AI QX remains neural driven year Tier the in year-on-year sequentially. vehicle partners, automotive. available from AutoPilot DRIVE was Systems self control. all computer. sequential $XXX at and level $XXX surround This ZF including the reached both year partially driver month commercially vehicles' safety, while cockpit driving The by growth significant networks Continental, ADAS the automated of two far AI offers that autonomous Last these announced Full

mass with XXXXs. plus cars. capabilities important This smaller be current for first for a slated centralized inside AutoPilot computing has the two takes Mercedes-Benz design market. year announced initial and of next October, autonomous also its functioning the Although, will system NVIDIA, high an as design and our in back pilots chosen generation computer. self-driving for cars AI into in milestone DRIVE dozens was is production early for and level This level it processors NVIDIA cockpit our two replaces robo-taxi wins. major vehicle win

the gross year was $XXX rest were million than year. the XX.X% XX% The million, XX%, Full sequentially $X.XX, up XX% QX margins year fiscal earlier. up with year and operating year. and $XXX charge market and $X.XX, QX in to EPS was P&L due lower dividends shareholders repurchases. the [port] and ago. expected and GAAP year-on-year, EPS to billion EPS balance from from of non-GAAP $X.XX and operating associated down was $XXX down to was share from primarily of was $X.XX, DRAM GAAP Non-GAAP our XX% sheet. year Moving down a GAAP were revenue respectively. a EPS GAAP XX% current the expenses other non-GAAP through XX% expenses and from and million combination We Full XX% quarterly $X.XX, components prior year-on-year non-GAAP conditions. up for prior was returned the year a

non-GAAP first buybacks. during last XXXX; and return or non-GAAP gross XXXX. With to we of and million be million; expenditures to of approximately minus excluding to We fiscal in to are approximately XX.X% or that, be the $XXX billion rates to operating and minus announced XX% expect quarter, and million we $X.X an capital are expenses turn As margins the be end of shareholders X%, the outlook $XX both quarter me non-GAAP be tax plus expected million of quarter and expected $XXX respectively. to are GAAP repurchased fiscal points; be items; $XXX expected fourth OI&E $XXX are revenue and GAAP form plus XX% through X%; the $XXX or to discrete respectively XXXX both XX million. for income be $X to expected and plan the plus billion let GAAP basis to minus are to dividends GAAP million expected fiscal we non-GAAP

XXXX, correction the the fiscal mark gaming. we expect in QX passed as For to bottom inventory we

details absence AI at to flat We this year for growth four the slightly expect for of are the CFO highlight events our available year. for selling Media Further on May be XX; with closing, our the to the financial commentary, Telecom information markets, like revenue other graphics, Stanley community; to presenting single the next call the in total results to OpEx We we on our Morgan of and I'd revenue to Web fiscal IR discuss most and and financial and in for is continue plan growth and In end will crypto down excess the areas in compensating grow site. to quarter included Technology, we take place upcoming the conference in our for high focus year, cars. February digits our self-driving financial invest and XXth. be of on XXXX earnings

questions? open call will Thank would We you for now you. Operator, the customers. please for poll

Operator

[Operator from Toshiya Hari comes the with our question Sachs. instructions] first of And line Goldman

Toshiya Hari

questions. two had I

to at you the saw and XXXX quarter. about guess weakness First, in about convince more XXXX your Jen-Hsun. in concerned ability is Are the I talked you gamer question Colette, the you out, this of can the all thing. incentivize and and on sheet. Colette, you. Colette a timing forward. expectations balance is it's the going inventory there as provide some up And if you pointed more was Thank question the [indiscernible] for color second just

Jen-Hsun Huang

it where to only the special that, we time early some didn't time high versions couple available When the very edition into boards came SKUs the we've XXXX, up the MSRP end. the was XXXX show in that launched months out first for new And ever And a versions. the the clocked short were addition launched generation after. were over marketplace the of and and after,

weren't for if the into mainstream understands you reasons And I We able all the the XXXX that conditions now. so weren't everybody with launch segment think to ideal, will.

the price ideal. much week games, and applying deliver technology to beautiful And the of for wasn't time new so both AI images of some now, just you the engines ray-tracing us to I new it's performance. out. came new just only one needed the games one starting technology; it just sacrificing think possible and that apply and And without to surely think right we with was RTX for going and it ray generation. games to to every without this But that it will is it to delivered remains are it simply work I Engine that the understand spectacular, the benefits out situation of are graphics that I take its with technology now with tracing and is finally When I takeaway RTX point games to on think ray-tracing each a one. that is future which agrees see the point card at AI next Metro I now to the more our single ray of graphics this new today. It that that such the the is the hoping more a the out had the and probably Unity stories image so and situ I think to in true starting performance at processing Battlefield then tracing technology is card ray-tracing performance look so excellent with the best every best biggest game were day tracing that from everybody And a technology, incorporate out that Unreal at is they're Exodus, are really is for apply of talking that think come come single with of we X, are schedules itself. tied it's price new are developers engine combination reviews coming week everybody and and games. using make available the it's of what possible And RTX technology people and about out last RTX box, realizing the the RTX demonstrate in think ray that people

and that. the all big really take about that's be so in that, games excited And a to will so future news of future able advantage I'm

Colette Kress

your inventory Toshiya our balance. regarding So to answer question second

expected DGX. Turing, QX to incorporated at related we the to due dollars QX. just don't the XXX right as any is and end primarily further approximately Volta within write-downs have inventory Inventory the expect now of of finish weaker QX. write-downs Our we rose current And to million than

Operator

Your Evercore. with Muse next question from is C.J.

C.J. Muse

cloud, a sure is a and the into year? got spending half. handful in that little key of that you're I speak how What within deals what as that, ramp I if key for commentary of I data across on in ramp delayed. milestones through the then center the guess regarding guess high compute, you in color the nice a that higher give in guidance looking seeing reaccelerate you guess a pause more drivers, you're and Can of 'XX seeing embedded TX? the enterprise terms you're go we can are bit performance I'm business see pretty I fiscal the to second And to guess

Jen-Hsun Huang

is in all it was be experienced was with satisfied it slowdown broad I last and We temporary. cloud computing every enterprise of The level while. not There just quarter. The Earth the of has not across geography. across think that shipped to the we've been providers needs a every is, what a cautiousness it's certainly of vertical, based. saw that customers service And

think will And return demand so return. will and that I the customers

introduced does, year-over-year. it the are, is our situation Our cloud one And where in world's TX the and watts. deep GPU inference universal with first it's centers better in at data that the XX NVIDIA best if We and in a solution you does everything learning all look unquestionably capability, is dramatically take we world. GPU

Service Providers that their we Internet it world, And and production models all deploy so porting We're it data production. could high fits center. so hyperscale into around into optimizing with networks engaged the every

center have growth the the to have the machine to easier world, scientific see four our actively we've be that to rapids, deep and learning; but addition it and Big new on; is third Data, people developed now it data technology. able in with we partnerships excitement the And analytics drivers computing; people systems inference some call enterprises have that our to data So that’s that for we're make called large we've developed for around learning we different and working rendering. of adopt pre-configure analytics data that around enterprises companies partnerships because we IT

our new grow four ways enterprise us to have we So for business.

so And back the and pause growing. to forward we're looking releases get to we'll when

Operator

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

Vivek Arya

had question. just and I clarification a a

issues. been number the essentially through exceed you an Pascal? that gross clarification, margins? of year, been run gamers the mix to there's And to On As trajectory? upward you so when that last then the saw has be gamers, on is there question Jen-Hsun, the should margins? you I sell for of and lot balance you could normalized still your levels, on are gaming of we sales margins. what trajectory confusion, will you from to growth rate they look if more of the Turing normalized how any Colette, think about see the that that do is back understand the And gross a can inventory? of get Thank gross And give demand you. that, us reassurance at as we there's sheet product the for impact buying over to a think business? been the part But macro

Colette Kress

from contributor the products on the gaming. as has also of mix our margin, absolute provided the which market there. see of margin us from QX, is as to how our platforms, based level data goes gross largest but product. gross guidance within our our a center, it good margin, and we'll our mix of of well for products really mix gross just our within The confidence We On

Jen-Hsun Huang

there fundamentals games, gamers of content. It it's in hardware. gaming of recent exception because lowers with like than to good better are free There's The popularity e-sports that's multiplayer the a more has the play been than to for of Games downloadable entry, ever. not the barrier changed, competitive ever. shift

lower Fortnight PUBG still Legend, entry is Legends to and see so recently is could Apex is excitement of around the barriers with that you but popular. League popular, And still the

it and a genre lot to it to social this so great happens to And requires attracts of your play it's game and play other people. players, a hardware, where it's more is happens have it games be friends with to you bunch want be much whole a with of because because you competitive, stickier, both social,

ever I before. And think so that gaming vibrant as is

you grew. If to and you described Colette underlying surely earlier year out that averaged that methodology a our before, business you gaming take it compare and the

compare at -- rate growing, you if also notebook that If exciting. that's is which our you pretty think the compare I gaming

number new gaming year last we before just because and notebook of of XX% our out with designs that think had it CES an architecture, Turing, that performing and year-over-year. the grew now mentioned came I notebook make invention of And we business at created called energy efficiency can time. you really also at wonderful the Turing of Max-Q and same that because the are that high the notebooks

are the is business. think I be so growth dynamics a the to going to continue same and And gaming

Operator

And from line John with your Credit Suisse. comes question next and Pitzer the

John Pitzer

in accelerate Colette, I am, throughout is for year off of gaming. seems doing inventory full for I to the that channel all us out revenue the What numbers? though gaming above the size see if to you am like you appreciate right? to that you’re expectation confidence normalized fiscal year gaming normalized what talked it to the gave What guide on ask you have first demand growth through gives you going quarter, is math your worked I hit about. I can that if trying level well these data And

Colette Kress

I’ll let finish question. that start I’ll and Jen-Hsun

response of our in drivers with Turing the and Jen-Hsun's growth of the well to believe as that be of that in still of out what we wait how will are we year. expected in along the is the plays great our architecture, gaming, key Turing both intact architecture, into the We'll drivers everything's the terms underlying reason as And we do the gaming. with but really see rest continue. of do the So from why the of green lines believe as growth terms notebook, will terms have head

Jen-Hsun Huang

I math that The didn't consider is isn't think your wrong. probably you part notebook.

notebook is quite business GeForce large. Our

John Pitzer

follow-up about inside the just in ASICs. the side. center more [ITCM] us volumes to center. get us TAM talked on And grow data meaningful help your always center. curious then competition anecdotal to data comfortable the had What quarter I’m what’s think There's the either in on ASICs? you’ve talk help of competitor GPUs marked guys a calendar just data is wanting year, of first here maybe a new differences evidence that GPU and more to between about own I architectural that as time and/or do the their applications loss Clearly, you give fourth not going last macro that and some share should can

Jen-Hsun Huang

see high don't performance in We them computing.

in And that the and where so serve. them learning, we and I met haven't -- high performance areas deep computing found in haven't we

data that really quite both be of high methods, well I going competitively, don't analytics, methods. unquestionably computing that it the those think as driven serve, picture data large. data so driven, And we algorithmic to it's future deep is methods, this machine learning, whether in are see is And as is the we segments I computational really the learning, But it. bigger performance think that highly segments market

think so fundamental I the not trend And has changed.

in to grow ways data four center. four have new We the growth drivers, new

lot there, one progress great. course doing We're making of TX is is first of The inference. a

quite with be TXs. to going we're successful think I

And can into any You for Tensor literally for it remote just you saying got use Core, has data and inference, could it computing. can you you graphics, use high generation XX watts to use training, hyperscale keep it for you fits use could center. it that second performance it for

second This analytics. is brand us. is a data The thing way for new

services. is using over course accelerated You financial performance retail Data we because the solve data to in must and and entire dynamics of the problem this because and architecture, of in know marketplaces that for And to there's analytics of the e-tail, CUDA approach really the been healthcare, important the year, predict people. never of last stack. data Big and flexibility reengineered And and

we work I you new servers market Turing, accelerate, render to can So can for way. There is Rendering accelerated regular in called hope a us, it's and really because updates and images finally driving are RAPIDS. that that of they're in on the farms on photo we of brand important world are basis. we That can give a render realistic is really, millions guys that getting every a industries. we've successful then giant because many CSPs, upgraded for really far they're in. sales to company's our single company where coverage our and lastly, on that's retail us And doesn't But allow regular reach And single easy basis. healthcare network partners are they're reach. And enterprises to company. with company world's us and and the every insurance of and been come

We have great with HP. partnerships

with most reason of together relationships with And company is vendors. IBM. -- can now developed the for see the you because of storage. And storage came big problems they data And systems install, require and opportunities optimized attuned. Dell see bring these of And great are we've prop with partnerships deal and a and that the have we're all great they Cisco just that high We that pre-configured great the into we we to success and created. systems seeing these that. up lot performance And create that

business data to about center different we're optimistic enthusiastic so and have our And about we grow it. I'm four ways it.

Operator

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

Tim Arcuri

the is assumed the QX clarify a I to had just guidance. fiscal for what wanted clarification. I Colette, First, mix

you through assuming help year? you the year billion full gaming roughly, data through And year? revenue the $X.X us To in flat center Thanks. are maybe sequentially. have where fiscal shape the So I and does How shape guidance, QX. get then billion, you are guess Can that you that both to to from add your my $X.X was. question

Colette Kress

make-up guidance provided quarter many for as feel that and the today Our market have of of out. see guidance is the how we we'll different across to options a our that types just rollout We of terms next pans in segment. confident

or of definitely flat first going of be of be slightly down course to time. the several this will up expectation the than second correct to year, half have the Likely, are half quarters. build stronger the we're the in the we you the at year and that next case that will full expectation to is our the With that year, that over some for

Jen-Hsun Huang

one drivers. in keep is to Tim, have mind growth the we that of four things

We have four growth businesses.

center is unquestionably larger business data footprint It's than that is Our growing. our ever.

is growing. business ProVis Our

One workstation three now rendering. Our grow. is has business to ways

our with a stack now The talked workstation finally second notebooks thin, are science notebooks third a have design using and entry same is know, the business. marketplace an as for workstation gaming our And And and delightfully business. workstations vehicles. going into data our automotive is cars. to level others And and make all And been to is win Turing about so of before. able scientists we foot mainstream autonomous that that growth gross in happened technologies in ideal into you first is to workstations self-driving the terms be software into customer the they're lastly, announce. a is then growth plus, And Volvo the into announced We've data this year of we're we has workstation. investing, first never two

self I going driving for to think year a good this as well. so And be cars is

core So four we have four growth our businesses, growing. are all businesses

Operator

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

Stacy Rasgon

get question to is this wanted I the mix, First, Colette.

down the sequentially, this driver they're primary gross in was mix And of the your correct to gross margin I I that if But QX, normalized the said up. was going inventory write quarter for that margins. know had be XX.X. You

down flat guiding XXX points you're to on it And so basis sequentially it's so revenues. XX,

the an take the the driver degradation, I something there's primary of or gross we aware do of on of? that's between is that that indicator margin mix businesses going that is if from should guidance So be else

Colette Kress

We'll start first gross with margin. that on question

You're correct gross every our margin mix primary still driver quarter. single is the of

overall reduced QX down. to correctly You've our or remove numbers inventory the write changed

These see within that plan a there gross at best time around also on we products So are data our you our business our look the we shipping by different influence. move margins and mix within when QX, we of also year. we'll and at as have have estimates is gaming But this our business, we center would through platforms. what that

Stacy Rasgon

you the two, Which one inter intra between mix, whether the those or gross sequentially degradation into or of the QX? biggest between mix those So think the driver do of within it's what the margin business is bigger businesses. business businesses

Colette Kress

I think inter the pieces. it's more on

Now the a well. total keep as in mix as in mind moved as percentage we within low gaming our there a terms then QX different had and next in of of there quarter very

Stacy Rasgon

that Is within, So mix that the think the bigger the you mix there? the inter. said, it's you think drive you between between is businesses businesses like

Colette Kress

intra, the as The mix the the both within of segment well between as segments.

Stacy Rasgon

could I've But my down and I've down Colette, data slightly to Jen-Hsun said, growing. ProVis It's got full like got and growing. flat year you and you center was For if have thinking of mentioned even at for context the really I follow-up, bit mean within year auto you all. the the I revenues hard you full revenues gaming the idea of the and mean, a I within growing centers data that envelope the you're within much guidance it get center that. given? data be down that growing about how are if to guidance

Jen-Hsun Huang

have the visibility. the relatively tailwind, going in think way. in year. have limited short-term, little bit think could we to we And it's good We this near-term, a fairly remain with a Well, of don't I

data fact you matter that The how lot business. we're of The to the now the working learning, computing to ways center deep grow just on. out. that things analytics, more we'll This so turns computing. dynamics machine related our And we our doesn't data new go. is that to for is And we related fundamental And world guidance see as have learning. and four It's we'll a needs change. year of it update is

I think our learning good is deep as position ever. as

to good the four Our grow; taking ways enterprise. scientific to computing And have we're we have inference, and stack we as ever. now we have that data is analytics new and learning, position entire machine we as have rendering

really think have the I utility best it fantastic. we the utilization, platform right we of and so strategy, have And the is the

go. tailwind, have little a it as we so we think And we’ll bit year with just a I and fairly of report could good

Operator

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

Joe Moore

are. sense us a other context than around vendors? you My a of center you doing AMD different their what data data again quite worth wanted on guys is in business ask bit some talked me. give and their I which within from competition biggest about quarter, million about they're graphics how just applications $XXX maybe server and But doing graphics north had competition they're to call see surprised center. being, coming which the is

Jen-Hsun Huang

really see we is business center don't and computing on data focused anybody. Our

the slowed majority The percent creeping it competitor ending not really the a just has it. the And clear. at good it's of of unfortunately, and vast technology CPUs along world's on data few a today. only it's and very primary CPUs. that's advance is the year starting enough. runs of Our That's center But of and

data good. about in a And people centers new to going And is accelerated think CapEx, have university, are computing. fairly about so to continue either approach. our think going quite really I or find I to enthusiastic they're increase to And position are

And that so I are say would the positions. those largely

about latest GPU you really the high a out quite comparing to think turn competitively I way. think that expectation competitor's didn't If was GPU, and our

-- we've energy much is think I established the efficiency better. Turing

performance. as NVIDIA’s of a to the the core I the result be Volta we've highest is power And of allows the the performance at that fourth established end alternative. tensor TX times architecture one think the same four our

going And rich broad energy generation in a great I the And are really great then so generation really those that the software engineering an having or OEM reach to utility really to there adopt the architecture I just can And our utility. more provider service its be a And and strong any cloud applications. because as is increase for could that to utility think to of the you a there's imagine. we've in of because is lastly, and architectural costs position advantages. advantage, our resulted and those that lot efficiency achieved benefit of greater, architecture, it's are the stack best our reduce an of out, think way

Operator

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

Matt Ramsay

couple just of questions. I have a

some And base the the upgrade do is DRAM. for might it rollout? installed been talk wonder you. to The pricing performance commodity mean first of amazing that about taking you of the delivering new anything and obviously volatile. some of with you any in one for on pre-announcement is steps business games, software something do to relying as the you if effect about and like features an the has in on those features Thank Jen-Hsun, Turing take it. can platform pricing about how talk talked then but to the I you write that time overall? results software some about that there have given Colette, really I down having secondly, the noted roadmap seemed to you're it might big on might

Jen-Hsun Huang

big much existing a better the all, in processor Without single per The instruction at big see last can to the than image much price point performing of and generation, you architecture point. what's the it's Turing improvement, our concurrent And generation The single operations in way. our marketplace. time. jump the of is architecture at is every first better is in clock of so Turing and just all architecture last cash it the floating First the games. ray-tracing same highest so every Turing a architecture the the from do GPU is available your than

It's But growth a we about big of rendering its game of the big Core we're that before second making about. spoken gives with talked remember Core technology Turing. all So Matt the Tensor Engine And opportunities And about are film; incorporate for part doing leap take we computer deal, know the just are work move computer that out, earlier that's generation of of it. the for measure and work that ray-tracing. games and working done several just first great we're Turing every is Turing. and just ray-tracing. going we games; us; the single graphics all existing for that And machine with games advantage we've forward boost developers It of advancing coming and it the for that for inference. lastly data and well that without will new that's notebooks; a to of touching advancing all of learning anything, is big very that incorporate we've for very, a the is Tensor Epic course, training, ray-tracing very, RTX with well then analytics talked our we've to as comes with already for work advancing to generation inference, you all with us photorealistic as industry capabilities worth Engine the next the done ray-tracing. RTX graphics for clear Unity both Unreal the very graphics is Turing

of Turing us. so the a is for big last for was And that's why year so deal And reasons us. one busy

As into we rendering, into games. into data clouds, into centers, put Turing workstations, into video

Turing by really is really we're for so gigantic it. And excited us architecturally, a leap

of analysis, us. turbulent think I all in QX overshadowed I the think home for run it a was Turing but final the

Colette Kress

question price. down your It to them regarding period. DRAM in answer historical over is to volatile And coming the great prices, definitely also yes, they have been see

gross us a yes, Over the that long-term, to from is margin perspective. beneficial

out gross we probably our look to later, be incorporate margin. able that to the we will as into So or quarters

Operator

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

Aaron Rakers

around your and China you're China on your the Turing? Turing that assuming rebounds not Has be invoked these help whether to pricing in can or interested also And around bit changes an solutions annual gaming the how company the segment is, the I'd strategy important is curious market platform us of the to the comments and is discussion inhibitor. prepared your then you you assumptions? frame in gaming in new pricing I'm Building of particularly any market, mentioned a

Jen-Hsun Huang

the we The of but reviews People was so inhibitor XXXX the success. launch point a biggest was to at is fantastic. the love on did price The segment. our us, First ability great a are CES. that range we is great. The for inhibitor all, mid pricing part, launch XXXX, big launch couldn't

we now stack way have from to all so And mid enthusiasts. the great the range

versions clocked high. many The over we special there were appeared of so the of launch, for which out better turned so many additions segments. and But it there the that other MSRP part, price because now all at were point time have is pricing our the

that's And going is this gaming very important China's an an market. so market terrific. And it's and important I to rebound. confidence

Operator

Lipacis the of from line Mark question next with Jefferies. your And comes

Mark Lipacis

fiscal -- where's deeper, Could you clarification. you of versus is OpEx? focus the then of a the that just OpEx would focus layer extent the investment, in repeat grow to higher SG&A? you And this you. Thank thought And maybe the Could what is this what one R&D year? you said what

Colette Kress

we indicated front repeat fiscal related to to off in I'll 'XX. our where in as us, we terms single-digits year over to the That cars. finished is opportunities of that start our gaming, see we AI plan self of high increase driving what well in transcript. OpEx in We as

in company. Our front are to focus heavy well strategies terms there as terms are both But in in significant high market of R&D as investments, of across obtain need board, in investments very us. what we of R&D a markets we go these the to

Operator

line RBC the Markets. Mitch And question your Capital from next with comes Steves

Mitch Steves

gross don't margin poke inventory. reached guys I if about profile, think the want almost crypto So the many But side to to XX%. too you and I due gaming holes downturn in on memory the

up if it So a get a even XX back make look long-term level? of months or it to let's more there to Xs Is you the I can't any why year target. call mid out reason guys

Colette Kress

On the gross for absolutely. allow us grow your long-term within definitely that. margin over margin to our a products mix yes And that question, our is continue drivers us have goal still the we to gross doing of there

entire focus we sell. that both value what the the of platforms cost added will we do, portfolio but of also So to components moving higher in the terms we

guidance still best look intact help that we'll we So over do. you see the quarter-to-quarter to that things give can And in all of place we can those absolutely long are the term, that. and to

Operator

And line Blayne question comes your the next with from Barclays. Curtis of

Blayne Curtis

Thanks. with particularly versus customers just Can any geographic maybe And documented. AI particularly still well run pretty with center it's data in number the or growth is area how how just and is perspective, pause to your hyperscale, widespread being curious perspective as curious slow and down? affected being memory us more AI a CPUs. Obviously, of this you greater perspective give rate

Jen-Hsun Huang

see dramatic. reason lot enterprises And the they we of for in it expect And We of just the to Blayne, the approach a us the but enterprise. use hyperscalars, the in an is still much pause use a for future. base is base smaller larger their most probably most activity that machine don't learning, much that graph stack called learning. It's things of today might a like because and data, be was decision they deep source might of business large spark use like created that. of run things and Databricks. open They or that, from trees, intelligence be running or amounts Berkeley or clustering, a applications regressions, they algorithms or like it data analytics, on top analytics for out on

use make of and route companies dinner that's and They do. do, you. do retail if take services riders inventory, the to best to recognize things, trying fraud companies matches there's healthcare what predict between detection, to which to drivers, try for deliver you -- companies and some and so food predicting And to it financial these

of applications those the today so big And analytics. learning data large use machine and developers most

world's built our RAPIDS partnering called with and architecture, IT process we TX. and called the the stack out we're in the of take stack to enterprise. So in And the the large invested to solution companies

their CSPs, we deep of using The enterprise with and and inference opportunity of pause And TensorRT. And growth increasing. so, work are the they I computation More do learning. we amount us. the TX end. expect them that for have to is fairly be more opportunity do with exciting a Meanwhile, will

a we've got to go. opportunities lot exciting so of And

Operator

line comes from Ferragu Street And Pierre your Research. with New next of question the

Pierre Ferragu

A Colette, last with position your question data to evolved year financial how XXXX. cloud for out centers I'm in figure year trying

about at XX%. I about up look cloud to the we into are Intel, XX%. CapEx, according look the up the cloud, if So at If going I CPU was

than CPUs? XXXX. wondering grow business, was Did And than have guys how much go up follow a have cloud you then did increasing I your you on revenues overall faster CapEx? within I faster So been you

Colette Kress

that on out start me Let piece.

start of growth half the As full many rate you the to our purchasing of latter the their can data in year. business hyperscales see for than our are as hyperscales from also we providers slowdown did talked about earlier terms on them the and more overall year. XX% of in cloud But grew call, our center

we still Now the CapEx are see percentage a very in we hyperscales. even without overall growth, providers that small cloud overall of strong the or

center the their focus likely focus of priorities one in very, of We data on AI, we're of the small necessary. as of a as and they is grow importance on businesses, of terms it. very to that But they they their are of percentage where top need still cloud compute areas in

of that So this the of half year, growth, we for slowdown growth year XX% track and with second in the phenomenal full year.

Pierre Ferragu

that's Is business But or your that cloud this all hyperscale that's grew even grew XX% overall well specifically with did that? what center growth, faster like your business. data

Colette Kress

on growth quarter the we in have hyperscale was quarter, depending Yes, And or it is the was. for a but what will that cloud year growth definitely opportunity not a earlier is mix the in what of in us fourth this. again

Pierre Ferragu

Do we like listen very be And to then When quickly and tuned be listen this And maybe others, Jen-Hsun seemed data to like to so Facebook seem very providers cautious players seem and center grow to on you spending we resume. in other outlook. their second they all confidence to when demonstrate half? on Google hear that overall second the year. XXXX, to say than the something you, they your they that all about do from should Is see? half, the you? seem more just What I've all you of they optimistic be clients bit around -- don't spending a you a because been

Jen-Hsun Huang

think down number. should we the growth is flat year, wait it maybe slightly to we're to guided enthusiastic you our application the we're solution biggest We We Pure, the about offer. have way the our we I takeaway And And And goes. about Well, second four market as how see position. drivers. for whole and the reach. We're also our probably just we've about do guided is enthusiastic for the whole about best expanded think half. and know, enthusiastic we it

with I feel even good are. position. feel our strategic about we guidance, our but pretty better we think where about I I concerned

look I to so And the forward through you year with guys. working

Operator

And your comes line of JP with last Morgan. question from the Sur Harlan

Harlan Sur

that it since approvals, China place to Chinese approve freeze economic government ban weakness, in first the This environment policy games. the on slower is last probably is we has related? gaming is know government half had or a year. it On recently I of new gaming Because started the the think new

weakness business, of given So just much the economic you how China is coming the China know gaming overall what from versus bans slower environment?

Jen-Hsun Huang

just don't know that, year. end I economy growth consumer towards of apart, the market of relatively the the economy. Harlan. could we the that is tease is the apart, in know the that slow analysis We final China, And China that tear

describing so and as seeing the And the so of done inventory well excited positioned. leave pass-times gaming and And bottom has recovering. we're their Colette the one All prospects we're culture we looking is important the of forward our there. really this to it us, that job most of near-term, I'm in about super a good and issue behind things we're leave

We is best of have in stack the full a architecture is The unquestionably Turing RTX. It the world. fantastic.

single now can great that We price point. the have we at best every performance notebooks market have the buy. And

status And so I'm with goes should be It guys a good reporting to as looking on. year. year our forward you the

Operator

any for now I'll to you. Thank back over closing remarks. Jen-Hsun the turn call

Jen-Hsun Huang

it record XXXX but was a finish. was a disappointing year

This channel inventory back issue quarter, track. to we on the behind and get expect put us

performance strong. ahead our is position and the accelerated of computing, in pioneer and autonomous enormous. the unique As computing, And high AI opportunities graphic, machines remain

growth these enthusiastic ever. as today. as Thanks everyone about are We joining us opportunities for

Operator

concludes today's conference And this call.

disconnect. now may You

NVIDIA (NVDA) 14 Feb 192019 Q4 Earnings call transcript