Q1 2021 Earnings Call

[music].

Good afternoon, My name is Josh and I will be a conference operator today at this time I would like to welcome everyone to Nvidias financial results Conference call.

All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If he would like to ask a question. During this time simply press star followed by the number one on your telephone keypad. If he would like to withdraw your question. That's the town Keith Thank you.

On the Gen. Koski you may begin your conference.

Thank you good afternoon, everyone and welcome to 10 videos conference call for the first quarter fiscal 2021 with me on the call today from India, or Jensen, Wong, President and Chief Executive Officer, and Colette, Kress Executive Vice President and Chief Financial Officer.

I'd like to remind you that are calling being webcast lies on and getting with Investor Relations website via webcast will be available for replay until the conference call to discuss our financial results for the second quarter fiscal 2021. The content of today's call is invidious property. It can do it produced or transcribed without prior written consent during.

This call we may make forward looking statements based on current expectations. These are subject to a number of significant risks and uncertainties and our actual results may vary materially for a discussion of factors that could affect her future financial results in business. Please refer to the disclosure in today's earnings release, our most recent form 10-K and 10-Q.

And the report that we May file on form 8-K, with the Securities and Exchange Commission. All our statements are made as of today May 20 for 2000 and training based on information currently available to us except as required by law, we assume no obligation to update any such statements.

During this call we will discuss non-GAAP financial measures. We can find a reconciliation of these non-GAAP financial measures to GAAP financial measures in our CFO commentary, which is posted on our website with that let me turn the call over to Jensen.

Thanks Simona.

College Grads are quarterly results.

Look to think those who are on this one lines of this crisis.

First responders healthcare workers service providers.

I would suppose everyday with their bravery unselfishness.

I also want to acknowledge the incredible efforts of our colleagues here in video.

Despite many challenges there barely broken strides during one of the busiest periods in our history.

Our efforts related to the virus or focused in three areas.

First we're taking care of our families and communities.

We pulled in raises like six months to put more money in our employees hands.

And then video and our people have donated thus far more than 10.

Second.

We are using a video is unique capabilities to fight to bars.

Great deal of science being done here technology for acceleration.

Some of them any examples including sequencing the bars.

Analyzing drug candidates.

Imaging the bars molecular resolution with quite electron microscopy.

And identifying elevate a body temperature, where they are cameras.

And third because cobot 19 won't be the last killer virus.

We need to be ready for the next outbreak.

And video technology is essential for this.

End to end computational defense system.

A system that can detect early.

They sold the development of the vaccine.

Contains a spread of disease and continuously test the monitor.

We're racing to deploy beyond video Clara computational health care platforms.

Ladies apparel bricks can accelerate genomics analysis from days to minutes.

Imaging will continue to partner with leading research institutes did devotes they'd be aren't the only models to detect infections.

And cleared Guardian will connect argue cameras and microphones and hospitals to help overloaded style watch over patients.

We completed the acquisition of Mellanox on April 27.

Mellanox is now and videos networking brand and business.

And we'll be reported as part of our data center market platform.

And Israel is now one of them videos major technology centers.

The new and video has a much larger footprint and data center computing.

End to end and full stack expertise and Davidson architectures.

And tremendous scale to accelerate innovation.

And video Mellanox are perfect combination and position us for the major forces shaping the I.T. industry today.

Data centers scale computing.

And ally.

From micro service cloud application, the machine learning and AI accelerated computing and high performance networking are critical to modern data centers.

Previously.

CPQ compute nodes, what's the unit of computing.

Going forward.

New unit of computing is an entire data center.

The basic computing elements are now storage servers.

If you servers and GPU servers and are composed and orchestrated by Hyperscale applications that are serving millions of users simultaneously.

Connecting these computing elements together is the high performance Mellanox networking.

This is the era of data centers scale computing.

And together and video Mellanox can architect.

End to end.

No no. This is an extraordinary company and I'm thrilled that we're now one force to invent the future together.

Now, let me turn the call over to collect.

Thanks Jensen.

Against the backdrop of the extraordinary events unfolding around the globe, we had a very strong quarter Q1 revenue was 3.08 billion up 39% year on year down 1% sequentially and slightly ahead of our outlook, reflecting upside in our datacenter and gaming platforms.

[music].

Starting with gaming revenue of 1.34 billion was up 27% year on year and down 10% sequentially.

We're pleased with these results, which exceeded expectations in the quarter marked by the unprecedented challenge of the Cobot 19, Let me give you some color.

Early in Q1 as the epidemic unfolded demand in China was impacted with I cafes closing for an extended period.

As the virus spread globally much of the world started working and learnings from home and gameplay surge.

Hopefully we have seen 50% rise in gaming.

Played on RG force platform, driven both by more people plane and more gameplay per user.

With many retail outlets closed demands for our products has shifted quite efficiently to E tail channels globally.

Yeah, I mean laptops revenue accelerated to its fastest year on year growth in six quarters, we're working with our Oems channel partners to meet the growing needs of the professionals and students engaged in working learning and plane at home.

In early April our global OEM partners announced a record new 100 in video GE forced Howard laptops with availability starting in Q1 and the most to SAP to ship in Q2.

Laptops or the first to use our high energy source Archie acts 2080, Super and 27 D Super Gpus, which have been available for desktop since last summer.

In addition, Oems are bringing to market laptops based on the Archie X 26, TGP you adjust $999 a price point that enables a larger audience to take advantage of the power and features of our checks, including its unique retracing and AI capabilities.

These launches are well timed as mobile remote computing need to accelerate.

The global rising gaming also lifted sales of in video intense don't switch and our cultural business.

Driving strong growth, both sequentially and year over year.

We collaborated with Microsoft emerging to bring our T X Ray tracing two minecraft the world's most popular game with over 100 million gamers monthly and over 100 billion total views on Youtube.

Minecraft with Archie looks astounding with realistic shadows in reflections light that reflects reflect added scattered through surfaces as naturalistic.

Effects like fog.

Reduce for our off the charts arch technical called it a job dropping starter and PC World said it was glorious to be hold.

Our Rts technology stands apart not only with our two year lead and retracing.

But with its use of AI to speed up and enhanced games using the Tensar core silicon on our Archie ex class GP is we introduced the next version of our AI algorithm <unk> deep learning Super sampling in real time do you Lss 2.0 can fill the missing Dutch from every frame.

Doubling performance it represents a major step function from the original at it and it can be trained on non gaming specific images, making it universal and easy to implement.

The value add momentum of our Archie ex Gpus continue to grow we have a significant upgrade opportunity over the next year with the rising tide of Archie ex enabled games, including major blockbusters like Minecraft Enfagrow problems.

Let me also touch on our games streaming service GFM, which exited beta this quarter. It gets gamers access to more than 650 games with another 1500 inline to get Onboarded. These include epic games fortnight, which is the most played game on GFN and other.

Popular titles, such as control Destiny to and league of legends with cyber power joining in the fall.

Launching in February GFN has added 2 million users around the world with both sign up and hours of game, playing boosted by stay at home measures.

Yes, and expand our market reach to the billions of gamers with underpowered devices. It is the most publisher friendly developer friendly games streaming service with the greatest number of gains and the only one such supports retracing.

Moving to pro visualization.

Revenue was $307 million up 15% year on year and down 7% sequentially year on year revenue growth accelerating Q1, driven by laptop workstations entering adoption.

Automotive three D. visualization software support in video R.T.X.G.P. It was.

This enables designers to take advantage of R.T.X. to produce more like life designed in a fraction of the time versus C.P.U. They systems over 45, leading creative and design applications now take advantage of R.T. Rex driving a sustained upgrade opportunity for padro power.

Systems, well also extending the reach.

We see strong demand and verticals, including healthcare media and entertainment and higher education among others.

Healthcare demand was fueled in part by covert 19 related research and Siemens, Oxford and caption Hell caption health received F.D.A. clearance for an update to would A.I. guided ultrasound, making it easier to perform diagnostics quality cardiac ultrasound.

And in media and entertainment demand increases companies like Disney deployed remote workforce initiatives.

Turning to automotive and robotic autonomous machines.

Automotive revenue was 155 million down seven per cent, you're near and down 5% sequentially.

Automotive industry is seen a significant impact from the pandemic and we expect that to affect our revenue in the second quarter as well likely declining about 40% from coupon.

Despite the near term challenges are important work continues we believe that every machine that moves someday, we'll have autonomy capabilities are.

During the quarter X. 10 introduced the piece seven an all electric sports sit down with innovative level three automated driving features powered by the end video drive Agey <unk> Xavier A.I. compute platform.

Are open programmable software defined platform enables x. time to run is proprietary software while also delivering over the air updates for new driving features and capabilities.

Production deliveries of the P. seven with in video dry begin next month.

R.M. care architecture Willpower, our next generation in video drive pots on called Orrin delivery more than six x. the performance of executor solutions and sport X. better probably efficiency with.

<unk> gave ability the drive pop form will extend from driverless rubber taxis, all the way down to N. windshield driver assistant.

Systems shipping just if you watch a power.

Customers appreciate the top to bottom platform all based on a single architecture letting them build one software to find platform for every vehicle and mostly.

Lastly in the area robotics, we announce that B.M.W. group has selected the new in video Isaac robotics platforms to automate their factories utilizing logistic robots built on advanced A.I. computing and visualization technologies.

Turning to data center.

Quarterly revenue was a record 1.14 billion.

80 per cent you're on your end up 18% sequentially crossing the 1 billion Mark for the first time.

Announce last week. The a 100 is the first <unk> architecture G.P.O., Although just announced a 100 is in full production contributed meaningful to Q1 revenue and demand is strong.

Overall data center demand was solid throughout the court or it was also broadbased across hyperscale and vertical industry customers as well as across workloads, including training in French and high performance computing, we continue to have solid visibility into two two.

The a 100 offers the largest leap in performance sedate over our eight generations of D. pews boosting performance by up to 20 x. over its predecessor. It is exceptionally versatile serving as a universal accelerated for the most important hyper forms workloads, including a I try.

<unk> infringe as well as data analytics.

Intific computing and cloud graphics.

Beyond its leap performance inverse fatality, the a 100 introduces new elastic computing technologies that make it possible to bring right size computing power to every job and most too much time instant G.P.U. capability allows each a 100 to be partitioned into as many.

Seven smaller G.P. two instances Conversely, multiple anyone hundreds interconnected by our third generation and the link can operate as one giant GPU forever larger training Tosh. This makes the a 100 ideal for both training and for in French.

A 100 will be deployed by the world, leading cloud service providers and system boulders, including Ali Baba Cloud Amazon Web services by do cloud Dell Technologies, Google Club top form H.P.E. and Microsoft does your among others. It is also getting adopt.

It by several supercomputing centres, including the National Energy Research Scientific Computing Center, and the Jewel Supercomputing Center in Germany and are gone National Laboratory.

We launched and shipped the D.G.X.A. 100, our third generation D.G.X. and the most advanced A.I. system in the world. The <unk>. A 100 is configurable from one to 56 independent Gpus to deliver elastic software defined data center infrastructure for the most.

Demanding workloads from A.I. training and infringe two data analytics.

We announced two products for <unk> <unk>, a 100 for larger commercial off the shelf servers and <unk> <unk>, an x. four micro edge servers supported by for a guy optimized cloud native and secure software the <unk>.

<unk> is built for A.I. computing at the edge.

<unk> hospitals retail stores farms in factories kind of securely carry out real time processing of the massive amounts of data streaming from trillions of edge sensors, and <unk> makes it possible to securely deployed and manage and updates fleets of servers remotely.

Ejects is also ideal for the massive computational challenge five G. networks, which we are are working on with our partners like Ericsson I'm Avenue.

Additionally, we announced queued up 11, and other important software harnessing B.A., one hundreds performance and University <unk> to accelerate three of the both complex and fast growing workloads recommendation systems confrontational, a guy and data science.

First Nvidia Merlin is a deep recommend date or application framework that enables developers to quickly build state of the art recommendation systems, leveraging our pre trained models with billions of users and trillions of items on the Internet deep recommendations or the critical engine powering virtual.

Every internet service.

Check it in video Jarvis is a G.P.U. accelerated application framework that makes it easy for developers to create deploy and run end to end real time conversational A.I. applications that understand terminology unique to each company and its customers using both vision and speech.

A man for these applications are surging added the shift to working from home Tele medicine and remote learning.

And third in the field data science and data analytics, we announce that we are bringing end to end G.P.U. acceleration too Apache spark and analytics engine for big data processing that uses more than 500000 data scientists worldwide.

Native G.P.U. acceleration for the entire spark pipeline from extracting transforming unloading the data to train him to in French delivers the performance and the scale needed to finally connect the potential a big data with the power of a guy.

Adobe has achieved a seven x. performance improvement any 90% cost savings in an initial tests using G.P.U. accelerated data analytics with park.

Celebrated computing platform continues to gain momentum underscored by the tremendous success of G.T.C. Diff digital.

Our annual G.P., a technology conference, which shifted the spring to an online form out more than 55000 online developers any I research registered for the online event, which includes hundreds of hours a free content from A.R. practitioners, an industry experts who leverage in videos plot farms or.

<unk> is now 1.8 million developer strong.

Times like these truly test computing platforms metal in the utility it brings to scientists racing for solutions researchers around the world are deploying R.G.P.U. computing platform in the fight against Coca 19.

Scientists are combining A.I. simulation to detect changes in pneumonia cases sequence the virus and seek effective bio molecular compound for a vaccine or treatment.

The first breakthrough came from researchers at the University of Texas, Austin, and National Institute of Hell, who use the G.P.R. accelerated application to create the first three <unk> three d. atomic scale map of the virus using and video Gpus.

This was followed by researchers at Oak Ridge National Laboratory, whose grade 8000 compounds to identify 77 promising drug targets using the world fastest Super computer summit, which is powered by more than 27000 in video G.P. is.

<unk> 100, G.P. news at Oakridge are in high demand as they can analyze 17 million compound protein combinations in a day to help understand the virus spread pattern the University of California, San Diego researchers ported there Microbiomes analysis offered to G.P. using the sandy.

Goes supercomputing cluster of 500, X. analysis beat up from what some people are more susceptible to the virus.

Okay moving to the rest of the P.N.L.

Q1 gap gross margins with 65.1% and non gap with 65.8% up see continually and year on year, primarily driven by G. force GPU product mix and higher data centres sales.

Two one gap operating expenses for 1.0, Threebillion and non gap operating expenses were 121 million up 10% and nine per cent you're on your respectively.

Q1 gap P.P.S. was $1.47 up 130% from a year earlier and non-GAAP U.P.S. was $1.80 up 105% from a year ago.

Two one cash flow from operations with 909 billion.

Before I turn to the outlook, let me make a few comments on our Mellanoxs acquisition.

Yeah, I'm, the strong strategic and cultural fit the Jensen, it's discuss Mellanoxs has exceptionally strong financial profiled the company reported revenue a 429 million in its march quarter accelerating to 40% you're on your gross with gap and nine gap gross margins in the mid to high.

60% range, we expect the acquisition to be immediately accredo to Nongaap gross margins Nongaap earnings per share and free cash flow.

We aim to retain the full mellanoxs team and accelerate investments in her combined road map as we jointly innovate on our shared vision for the future of accelerated computing.

With that let me turn to the outlook of the second quarter, a fiscal 2021, which includes a full quarter contribution from Mellanoxs, we've assumed in our outlook the potential ongoing impact from Coven 19, we expect our automotive platform sales to be down 40.

Per cent sequential basis, and prose is to decline sequentially.

In gaming, while we will likely see ongoing impact from a partial operations are closures of I cafes, and retail stores, we expect that to be largely offset by shift to e. tail channels overall, the precise magnitude of the impact is difficult to predict given uncertainties around.

Reopening of the economy.

Overall, we expect second quarter revenue to be 3.65 billion, plus or minus 2%. The contribution of Mellanoxs revenue is likely to be in limbo teams percentage range of our total two two revenue we are providing this break out to help with compatibility between q. wanting to.

Too, but going Frodo become an integrated part of our data center markup toxin.

Gap and non-GAAP gross margins are expected to be 58.6% and 66%, respectively, plus or minus 50 basis points. They sequential decline in gap gross margins, primarily reflects an increase an acquisition related costs <unk> most of which.

Or nonreoccurring.

<unk> operating expenses are expected to be approximately 1.52 billion and 1.04 billion respectively.

<unk> change in gap operating expenses reflects any increase in stock based compensation compensation and acquisition related costs.

Happened down gap operating expenses for the full year are expected to be approximately 5.7 billion and 4.1 billion respectively for the full year stock based composition compensation and acquisition related costs also influence.

<unk> Oh I need are both expected to be an increase of approximately 15 million and 45 million, respectively gap and non-GAAP tax rates are both expected to be 9% plus or minus 1% excluding discrete items capital expenditures are expected to be approximately 200.

And 25 million to 250 million.

Further financial details are included in the C.F., a commentary and other information available on our I our website new this quarter. We have also posted an investor <unk> presentation summarizing our results and key highlights.

In closing, let me highlight upcoming events for the financial community next Thursday may 28th we will webcast, a presentation and q. and ate with Jensen on a recent product announcement moderated by ever core we will also be cowlings T.M.T. conference on May 27th.

Morgan Stanley's cloud secular winters conference on June 1st be evade technology Conference on June 2nd Needham's fourth automotive Technology Conference on June 3rd and NASDAQ Investor Conference on June 16th.

Operator, we will now open for question can you. Please post of questions place.

Certainly at this time I would like to remind everyone in order to ask the question Crestar than the number one on your telephone keypad and.

When you're first question concept and rakers with Wells Fargo. Please go ahead.

Yeah, Thanks, and congratulations on the on the solid quarter <unk> I'm curious is your commentary round visibility in the data on our side that that's the you know comments over the last couple of quarters, how how would you characterize your visibility today relative to maybe what it was.

Last quarter, and and and how do we think about the visibility into context <unk>, maybe you know into the back out of the calendar year they get.

[noise] things well for the question you are correct, we have into quarters ago that we were starting to see improved.

[noise] of the digestion periods in the prior overall fiscal year as we move into Q. too, we still have visibility and solid visibility into our.

For overall data centers. So this time I'd say they are relatively about the same of what we had seen going into the Q1 period.

And we think that is a true indication.

About our platform and most particularly our excitement regarding a 100 and that's launch.

[noise] now regarding the second half of the or.

Seen broadbased grows in both the Hyperscale and the vertical.

Record levels in our <unk>.

Terms of infringing continuing to grow as well as well as we're also expanding into into the edge a guy.

Are strong demand.

Hundred products, including the Delta Board, but also in terms of our D.G. action. So as I was just starting a initial route.

However, we do got.

So it's still a little bit too early for us to give a true certainty in terms of the macro situation lots in front of us, but again, we feel very good about the demand for a 100.

That's good with Bernstein Research. Please go ahead.

Oh nice things were taking my questions. The first one of the follow up on your gaming commentaries for the mentioned a couple of off that's coded potentially still ahead. When you tell a tail end and maybe also to each other [laughter].

Pleading gaining was kind of why they seem to Cuba. They know what how the typical seasonal crowder.

So which is typically.

There is and what are the kinds of things, we should be thinking about when it comes to seasonality collect in the in the key to run the business button.

<unk>, let me start and all she jumps and also wants to out onto it I think you're talking about our sequential between <unk>.

Right.

Right some of the pieces that we had seen related to cope with 19 and Q1 may carry.

How about 19 in fact.

Had an impact in terms of our retail <unk> <unk>. However, as we discussed it efficiently moved to overall etailz, we have normally been seasonally down in desktop between coupon in Q. too.

But we do see the strength in terms of laptops and overall councils as we move for Q1 222.

Gross sequentially between.

For overall gaming business I'll turn it over to jumped into C.P. out additional commentary.

No goes good.

<unk>.

No I I I guess just both.

In prior years, we've seen it grow like very strong double digits, obviously, the mixed with a business was different back then.

But do you think that the the kind of I remember thinking you know kind of it's up somewhat you don't is there any chance that it could be applied gun, but what we've seen in terms of like typical higher level than the power like can you give us any something on the two that would be really don't.

When we think about that sequential growth, we'll probably be it and the low moving up to probably the mid to single digits in terms of that's what our guidance is right now and we'll just have to see how they're quarter goes.

Yeah, I guess, that's very helpful blood.

They see the thing that I would add to this I would say.

The guidance is exactly what what what color mention but if you look at look at the Big picture.

There's a few dynamics that are working really well no Subaru person.

<unk> Oh man.

Was phenomenal.

<unk>.

We're shipping just about every game developers went on to our T.X. Ray tracing.

Yeah, I think it's it's a foregone conclusion that this this is the next generation because his way computer graphics going to be in a few true. So I I think our taxes on the second.

The the notebooks that we've created with artists and that's cute.

Just doing great and then we got 100 notebooks and gain books design for either mobile workstations or what we call in video studio for designers and creators.

And and the tongue was just perfect.

You know with with with everybody needing to stay at home.

Mobile.

Gaming platform in a mobile workstation.

What's your perfect time, and then of course.

Yeah, you you guys know quite well that that on Nintendo switches to.

Or there are three <unk> three games in the world.

The top news in the World today are a fortnight.

Minecraft.

And animal crossing.

All three games on video and.

That's right.

We have old dynamics, working and also rubber and then we'll just go to we've just got to see how it comes up.

But if that's helpful.

You know things.

You know next question comes from Joe more with Morgan Stanley. Please go ahead.

[noise] great. Thank you I want to ask about the roll out of of Amp here.

How quickly does that role into the various segments between hyperscale.

As well as on the D. Jack side as well as on H.B.C. side, and you know if it is it it's moved trends and is there you know where I remember when you were launched both there was can you know how you see that ramping up with the different customer segments.

Yeah <unk>.

So so first month against about.

Accelerated computing is Mel commonsense and do centers.

There wasn't a good case, when we first lunch volcker.

Yeah, we went back to Volta Volta was the first Jenna.

Training in a really serious way and it was really focus on training those books been training and <unk>, we didn't come until later, where the inference version called T. for.

But over the course of the last five years, we've been exists already workloads that are now diversifying in data centers. If you take a look at most of the most of the Hyperscalers machine learning is now pervasive deep learning is now pervasive.

The the the notion of exists already be burning them for learning.

Using our G.P. uses there's no common sense then used to be.

You know people people People's still saw it as a something something esoteric, but today data centers all all over the world expect a very significant part of their data center being exposed to what you to use.

The number of workload that we've accelerated sense in the last five years six spanned a tremendously.

You know, whether it's whether it's imaging or video or conversational <unk>, a deep recommendation deep recommender system, so that like probably unquestionably at this point the most important used to learning motto in the world. It's just so that the number of applications. We now accelerated was quite diverse.

And so that's that's really that's contributed greatly to the ramp of Amp here. When we when we started to introduce rent M.P. or to to the data centers. It was very common sense sensical to them that they would they would adopted.

Large amount of work load that's already <unk> immediate you to use and as you know, our our G.P. as or or Arkansas were four compatible with backwards compatible everything that runs on T. four runs on a 100 everything that runs on the 100 runs on the 100 and.

So I think the transition is going to be really really smooth on the other hand.

Because the 100 and keyboard, which by the way do you want her and T. forehead, a headache recorder there was sequentially up.

And then on top of that we we outgrew with the a 100 shipment.

They 100 and.

The 100 and T. for our now are now quite broadly adopted and even in Hyperscalers for the A.I. services.

In cloud computing.

And and vertical industries as a court mentioned earlier, which is which is almost roughly half rubber overall H.B.C. business.

All the way out to the edge, which had a great quarter.

And then and then a much smaller part of course, you know supercomputing is important.

But it's it's a very small part of a high performance computing, but that's also that's also we also shift the 100 to Super computer centers.

So I I think maybe the the the general census visit the summary of it is that the number work load squeaks already to computing as continued to grow.

The adoption machine learning you know in all the clouds and.

Hyperscalers has grown.

Oh, the common sense of using acceleration.

There's no foregone conclusion.

And and so I think we're we're we're wrapping into into a a very receptive market with a really fantastic more from where the really fantastic product.

<unk>.

<unk>.

So next question comes from <unk> are you up with Bank of America. Please go ahead.

[noise]. Thanks to take my question and congratulations on the strong growth then execution, just a quick magnification collette, 66% kind of a new baseline for gross margin and then the question Jensen for you is give us a sense for how much inference as a work flow.

As a product are expected to contribute I'm, just curious where you are in terms of growing in the infringe an idea I market than van R.V. kind of an attorney off a at their penetration.

Yeah.

20 start on the first question regarding the gross margin and or gross margin as we look into Q. too. We are guiding cue to non-GAAP gross margin such 66% dishes would be another record gross margin corridor just as we.

Finished a overall record level, even as we are continuing right now to ramp our overall <unk> architecture with them not the queue to also incorporates mellanoxs mellanoxs is have very similar overall margins to our overall data center margins as well.

But we see this new baseline.

A great transition and likely to see some changes as we go forward. However, it's still a little early to see where are these gross margins logo on but we're very pleased with the overall guidance right now 66 for kids to.

Accelerated computing is just at the beginning of its journey.

If you look at if you look at.

But I would I would characterize it as as several segments first.

Is hyperscaler.

Micro services, which is all the all the services that we enjoyed today that has it on you know when when you shop on the web it recommends recommends a product. So when you when you're watching a movie recommend some movie recommends a song all of those or recommends news or recommends a friend or recommends a website.

The first 10 website.

Don't they recommend all of the all of the Recommenders that are powering the internet <unk>. The reason why do collecting so much data the more data they can collect all the more they could predict your preference.

End up predicting your preference is the core to the personalized mm.

Used to be used to be mostly based on a C.P.U. approaches are going forward. It's all based on deep running approaches.

Results or or much more superior and and a few percentage change in persons prediction accuracy.

Result in tens of billions of dollars worth of comics and so this is very very big deal and the the a shift towards deep learning and Hyperscale micro services Microservices is the still ramping second this club.

You know cloud cloud is 100 billion dollar market segment or by T. today growing about 40% into a Tolkien dollar opportunity. This cloud computing as a single single largest.

I cheat industry transformation.

The two powers that that is really the forced the two courses that is really driving produced in a business is a.

And cloud computing.

You know we're perfectly perfectly positioned.

To benefit from these two.

Powerful forces. So the second is this club computing and that journey is is has has a long ways to go then the food.

His industrial edge.

But in the future today is not that it's not the case today, but the the combination of I.O.T.

By G. industrial by June.

And artificial intelligence.

It's going to be is going to turn every single industry into a tech industry.

And whether it's in whether it's.

You know logistics or warehousing or manufacturing or you know farming.

Books, and industrial every single industry will become the tech industry and and there's there'll be trillions of sensors there'll be connected too little micro data centers and those data centers will be a millions we will be distributed all over the edge and.

And that that journey has just barely started we announce three.

Very important partners in three domain.

And and they're there to lead partners that that we felt that people would know, but we have we have several hundred partly sort of working with us on edge.

We announced Walmart poor smart retail, we announced the U.S. postal service the world's largest.

Male so wedding service and logistics service and then we announced last quarter B.M.W., who was working with us to transform their their factory into a a robotics automated stuck to the future.

So these three these three applications or or great. Examples of the next phase of artificial intelligence and we're we're amperes going to rent rent into and that is just really I was at its early early stages. So I think I think it's fair to say that we're we're we're really well position.

The two fundamental forces from I.T. today data center, a scale computing and artificial intelligence and the segments that it's going to is going to I'll make a real impact on all gigantic markets Oh, hyperscaler crowd and it says you know.

Like.

Thanks.

You know next question comes from C.J. Meuth with Evercore. Please go ahead.

They're good afternoon. Thank you were taking the question I guess, if I go to us to <unk> can you help us with what you think the growth rates are familiar likes to look like encounter 20, and then Jensen <unk>. It bigger picture question for you and and really not specific to healthcare more broadbased, but how do you think about.

The long lasting impact of <unk> on worldwide demand for a.

<unk>.

C.J. can you help me you cut out in the middle of your sons to me can you repeat the first part of it for me. Thank you Oh, sorry about that I'm curious if you could provide or you're a little hand, holding on on what we should think about for growth familiar Alex encountered 20.

I met a little early for us and as you know, we generally just give one quarter out and we're excited to bring the Mellanoxs team on board should we can start beginning the future of building products together for the overall market again, you've seen their overall.

Performance over the last couple of quarters. They had a great last year, they had a great Marge quarter as well and we're just going to have to stay tuned to see equally with them what the second half of the or look for them. Okay.

Yeah. So you're thanks for question <unk> the.

Right.

You know this pandemic is is really quite tragic and and.

All right and this reshaping industries and markets <unk> and I think it's going to be structural I think it's going to remain.

And I think your your question is is is really good because because.

Now's a good time to think about think about a where to double down. There's a few areas that I believe are going to be struck totally changed.

And then I didn't and I think that once I say there'll be very sensible.

The first is that the the world's enterprise digital transformation and moving to the cloud that it's going to <unk>.

Every single company can afford.

To rely just on on Prim I.T. they have to that has to be much more resilient and and having a hybrid cloud computing infrastructure is is going to provide them to resumes and.

And so so that's that's one and when when the world moves and accelerates a into this one trillion dollar I.T. infrastructure.

Transformation, which is now $100 billion into that journey, you know, it's growing 40% a year I I wouldn't be surprised to see that that it's already so cloud computing, you always going to it's only because of that.

Okay.

Oh it is the importance of creating a computational defense with them you know the the the defense systems are the most nations today are based on.

Radar, <unk> and and yet in the future in the future. Our defense systems are going to detect things that are that are unsuitable going to be infectious disease.

And I think every nation and government and sign scientific lab is no gearing up to think about what does it take to create a national defense system for each country that is based on computational methods and you don't have videos and it's already computer company, we we take something that that otherwise would take a year.

Cases in the case of Oakridge and they filter a billion compounds in a day.

And that that's what you need to do you need to find a way to.

<unk> annex already computational defense system that allows you to find inside inside does you know detects.

Detect early warning M.P. and then and then of course, the the computational system has to go through the entire range from mitigation to containment too you know living within a monitoring.

And so so a scientific labs are gonna be gearing up national lunch and then be during.

The the third part is that you I am robotics.

We're gonna we're gonna have to have the ability.

To be able to do our work remotely.

Video has a lot of robots that that are helping us in our lives.

And and without those robots, hoping within our labs, you know we'd have a hard time getting a worked on and.

So so we need to have we use have remote autonomous capability for.

To to handle all of these either dangerous circumstances to disinfect environments, the fumigate environments.

<unk> to clean environments to be able to to interact with with people were as little as possible.

In the event of of an outbreak all kinds of all kinds of robotics applications are being dreamed up right now to help society for.

In the case of in the case of another outbreak.

And then lastly.

I think I think more and more people are going to work permanently from home.

There's a there's a there's a there's a strong movement.

As a sort of companies that are that are going to court. They larger percentage of people working from home and then when people work from home.

But it's going to it's going to clearly increase the single the single best.

Entertainment, which is video games.

He'll games is going to represent a much larger segment of the overall entertainment budget of society.

And and so these are these are some of the twins I would say I would say on cloud computing they are.

I would say national labs.

Computational defenses from robotics, and working from home or structural changes that that.

Are gonna be here just to send these these dynamics are are really good for us.

You know next question comes from because she a hurry with Goldman Sachs. Please go ahead.

Hi, good afternoon, and thank you very much for taking the question I had one for Collette and then one for junk that as well if I may clutter I wanted to come back to the gross margin question. Your your Guardian July essentially flat sequentially, despite what I'm definitely better mixed with <unk>.

In an automotive got down 40% of sequentially.

That's the question is what are some of the offsets that are pulling down gross margins the current quarter and sort of related to that house would be thinking about the kittens and I'll pucks going forward given the six month, Poland, but you guys talked about on on the compensation side.

Then one quick one for for Johnson.

I was hoping you could comment on the current trade landscape between the U.S. from China, I feel like you guys shouldn't be impacted and and material way directly or indirectly.

Indirectly, but at the same time given the critical role you play in scientific computing I can sort of see a scenario, where some people may no claims that you guys contribute to to efforts outside of the U.S. So you can try to speak on that speak to that there'll be helpful. Thank you.

Thanks to share for your question so regarding our gross margins in the second quarter. Our second quarter Guide 66 is up sequentially from even a record level in terms of what we had into terms of coupon <unk> next record that we hope to achieve.

With our overall guidance is even with including our overall <unk> architecture. So typically when we transition to a new architectures margins can somewhat be a little bit lower on the on time, but tend to kind of move up and trend up over time.

Additionally, as you articulated our automotive is lower but also we're gonna see growth in some of our platforms in gaming such as councils, which may offset those too, but overall, there's nothing structural to really high my other than our mix in business and a round.

Amp here and it's transition.

Let's see the <unk> the the trade tension we've been living in this environment for some time this year.

And as you know the trade center and.

Has been has been in the background for coming up on a here a couple of weeks button longer and.

And.

In China's high performance computing systems are largely based on on Chinese electronics anyhow, and so that's I I think I think our our or condition won't won't material would change.

Going forward.

So two shit, let me let me respond to your second question that you had from me, which was regarding to our <unk> and our decision to pull for word our overall uncle into cute too, but this is something that we normally done later in the year we've.

It was prudent during the current coated 19, although our employees are quite safe. We just wanted to make sure that their family members also we're safe and have the opportunity to have cash up front. It is about a couple of months about four months earlier than normal and it is incorporated in.

Guidance for Q. too.

You know next question come from Berkeley, I guess with Jeffrey Please go ahead.

Hi, Thanks for taking my question question coming back to the eight 100 I'm trying to understand how this.

Kind of fits into the evolution of your your solutions said over time and the evolution of the demand for the applications is you know and I guess, what I think about going back. He you know you you had a solution, which is largely training base and then you you kinda introduce solutions that were were targeted more inference scene and now you have a salute.

Isn't it it sounds to my understanding that that is it it solves both in Francine and.

Training efficiently and so I guess I'm wondering is you know three years five years 10 years down. The line is is this part of the kind of general purpose computing or or acceleration framework that you had talked about in the past tense and we're we're amp here as it is kinda like an M.P. or class products.

Or is this you know would you still should we still expect to see it in front seems specific solutions in the market and then training specific solutions and then an M.M. pier solution for it but you know different class application. If you could provide a a framework for thinking about.

M.P. or and and and and those contacts I think that would be helpful. Thank you.

Yeah. Thanks, a lot more the the good question I think the.

If you take a step back.

Currently in our data centers. The current set up in data centers starting from from you know probably all the way back.

Six seven years ago, but really celebrating in the last five years and and then release already in the last couple of years.

We learned our way into it you know there there are three the three classes of workloads.

And they kind of came into big celebration overtime. The first class of workload that that we discover.

Was the the major work load was deep learning between.

Keep learning screen and and the <unk> the ideal the ideal set up for that today prior to N.P.O. or yesterday <unk> is the B. 100, S.X.M. with envy link H.G.P.U.'s on one board and.

And that architectures cost scale up it's like a super computer architectures like as soon as if the weather's simulation architectures. It you you were trying to build the largest possible computing node you can for one operating system cost scale up.

And the second the second thing that that we learn along the way was a then cloud computing started to.

Grow because researchers around the world needed to get access to.

They they celebrated their machine learning algorithms and and because they they have they have different degrees of budget and they they want to get into it a little bit more lightly and have the ability to scale up to load the nodes.

The perfect. The perfect model for that was actually a b. 100, P.C. I express not exercising but P.C. I Express that allows you to.

Offer one G.P.U. all the way up to many G.P.S.

And so that versatility be 100, P.C.R. express I'm not as not as scale a bowl in performance as the B. 100 ethics <unk>, but it was much more flexible for rentals renting was was really quite a deal.

And then and then and then we started to get into inference, and and you know where when our seventh generation a pencil R.T.S.R.T. seven.

Along the way we've been we've been able to solve it more and more and today, we we larger leaks all read every deep learning a infringe computational graph that's out there and and the ideal the ideal G.P. you for that was something that has the the reduce the precision.

Which is called a bit <unk> reduced <unk> precision.

I'm not with electronics.

That is focused more for for inference.

And and because airplanes is a scale out application, where you have millions of queries and each one of the queries are quite small.

<unk>, where you have one training job and that 120 jobs running per day.

It can be running for days and sometimes even weeks.

And so scale up application is is for one user that uses it for a long period of time on a very large machine.

<unk>.

It's four millions of users.

One of them have a very small queries in the query could left hundreds of milliseconds, where ideally you'd like to get done hundreds of milliseconds.

So.

The noticed I've got three different architecture, and then <unk> data center today, most data centers today has a storage server has C.P.U. servers and it has scale up exhibition servers with.

<unk> has scale out.

Servers with <unk>, and then have scale cloud computing flexible stover based on the 100 and and so.

The ability to predict workload is so hard and therefore the utilization of these these these systems will be will be will be spiky.

And so we created we create an an architecture is that allows for three things the things the the three characteristics of the M.P. or our number one.

It is the greatest generational leap in history, I mean, I I don't remember a generation, where we increased throughput for training and inference by 20 X., It's just a gigantic.

Retraining and for him friends. It is the Gigantically for the second.

The first architecture that is unified.

We could use this computational <unk> with the the the computation engine of Amp here <unk>.

Accelerate.

The moment the data comes into the Davidson from data processing, it's called P.T.O.. The the engine that which many of you probably know it's the the single most important computational engine in the world today for Big data.

<unk>.

Spark is used to all of the World 60000 customers. We finally have the ability to accelerate that and then it's amperes also good for training deep learning machine learning extra boost as well as deep learning all the way out the inference.

And so we now have a unified.

Acceleration.

Platform for the entire workload and then the third thing is it's the first two p. you ever the first six television platform <unk> last it.

You could reconfigure it you can configure it for either scale up or you can configure it for scale out.

When you configure it for scale up you're getting a whole bunch of G.P.S. together using M.B. link and create this one gigantic G.P.S. when you want to scale. It out that same computation node becomes 56 small G.P. is just.

He's one of those.

You six partitions each one is more powerful than Bolton.

I mean, it's really quite extraordinary.

And so so M.P.R.M. periods a breakthrough on on all of these parents you know that were poor performance, where the fact of a unified the workload and you can now have one acceleration cluster.

And and and number three it's a lasting you know you you you could use them in the crowd you could use it for instance, you could use it for training and so the versatility of Amp. Here is is the thing that I'm. Most excited about and now now you could have one $6 from cluster that that's served all of your needs.

Thank you Sir helpful.

Yeah, Thanks, a lot more.

Your next question comes from Timothy or carry with U.B.S. Please go ahead.

[noise]. Thanks, a lot actually I had to I guess Jensen first for you just on the data center business things have been very strong recently, you know obviously, there's always concerns the customers or you know pulling in cat backs, but it sounds like you a pretty good visibility into July.

But I guess last time.

You know <unk>.

They should really was so low that you would be immune to any digestion, but that wasn't the case. So I guess I'm wondering if you know things are different.

But I, but my question is how you handicap your ability to this time, maybe get through any digestion on the <unk> you know cat backside and then I guess second question Collette <unk> had been running like 220 a quarter.

And the guidance implies that it goes to like 460 a quarter so it.

Set up a lot is that all executive attention and is that sort of the right level was you look into 2021 sex.

Collect did you want to handle that person then I'll I'll do the.

Sure.

Oh, So let me so that might help you on the overall gap adjustments are the delta between our <unk>. If you look at it for the full year and what we guided we've probably have about 1.1, 0.5 5 billion associated with gap level expense.

Keep in mind, there is more in there than just our stock based compensation. We have also incorporated the accounting that we will do for the overall mellanoxs and it really good portion of those costs are associated with the amortization of intangibles and also.

In terms of acquisition related costs and D.L.C.. So our stock based compensation includes what we need for in video and also the Onboarding Mellanoxs. There is some retention with the overall onboarding of Mellanoxs, but for the most part it is just more kingdom and.

To the year for three quarters, which is influencing the stock based compensation.

Attend their civil there are several differences between our conditions, then and our condition today.

So the first the first difference.

The diversity of workload, we now like salary.

You know back then that then we were early in the inference, we're still early in our in Prince and most of most of the data center.

Originally you use would be fun.

And so <unk>. So today today, the versatility stems from data processing.

Do boarding and the number of the number of different types of types of models as being trained for deep learning is is growing tremendously.

From from detecting from training videos from training a model <unk> detecting detecting unsafe video.

The natural language understanding conversational, a nice to know a gigantic movement towards the Recommenders systems.

And so the number of different models that are being trained is growing the size of the models are gigantic recommendations systems or gigantic.

Their their training on models that are hundreds terror the data sizes hundreds of Terrabyte.

<unk> <unk>.

And it was it would take you know tens of hundreds of servers to hold all of the data that that is needed to train. These recommenders systems.

And so so the the diversity of urban from data analytics to training all the different models to the influence of all the different models, we didn't we didn't imprints.

Recurrent neural nets at a time, which was which is probably the most important modeled today text <unk> language model speech models are all recurrent neural net models and so those models.

Work early for us and at the time. So number one is the diversity of workload. The second the second as the installation of to cloud computing.

I think that that it's celebrated cloud computing is a movement that that is going to be a you know going to be a multi year.

Not a decade mom a transition.

From where we are today, it's only 100 billion dollar industry segment of the I.T. on the screen will be a trillion dollars someday and and that the movement is just starting we're also much more diversified out of the clouds.

You know at the time.

<unk>.

Largely were already solely from went for a deep learning and today Hyperscale only represents about house.

So.

We've diversified significantly.

Out of out of cloud not out of <unk> club, including vertical industries and a lot of that has to do with a G.I.M. friends and and as I mentioned earlier with you know, we're working with with Walmart and B.M.W., and U.S.P.S. and and and that's just the tip the iceberg.

So.

So I I think the the conditions are are a little different and now what I would say last week is is I have here I mean, we we are we rent but.

A few weeks.

Even though it was quite significant there was a great round the demand is is.

Topic. It is the best ramp we've ever had the demand is is the strongest we've ever having data centres.

And.

And we're we're starting to ramp of a multi year round yeah. So.

So those are those are some of the differences I think the conditions are very different.

<unk>.

Yeah things about him.

You know next question comes from Heartland served with J.P. Morgan. Please go ahead.

Afternoon, Thanks for taking my question.

You know the today show the importance of networking networking fabric and another one last acquisition wait for example, when you guys movies from.

So did you x. one to vote to D.X. two days they didn't change the G.P. chip said, but by adding a custom networking fabric shifting more mellow maps network interface cards among other things.

I still have a pretty significant improvement in performance per G.P.U., but now when we think about scaling out compute acceleration to data centres skilled implementation I don't smell relapses eastern it's switching platforms differ from those provided by other large networking Oh, yeah I'm somewhat phone has been your long term partners and then.

How does that Shoeless acquisition fit into the switching a networking strategy as well.

Yeah, great. Thanks, a lot Harlan appreciate the question so so.

DGX you know just our third generation digits.

And and it's really successful people love it it's the most advanced instrumental in the world.

If you're serious yeah researcher this is your instrument.

<unk> in the D.G.N. there are eight.

A 100.

And there are nine mellanoxs mix, the highest speed mixed anyhow.

And so so we we have a great appreciation for for Highpoint networking high performance not working on Harpercollins computing go hand in hand, and the reason for that is because the problems. We're trying to solve no longer fit in one computer no matter how big is.

And so has to be distributed and when you distribute they'd computational workload such intense scale. The <unk>. The communications overhead becomes one of its greatest bottlenecks, which is the reason why why no no. It's just so valuable there's there's a reason why those companies, so precious and and really really a jewel and the one of.

One of a <unk>.

And so so <unk> and it's not just about the the links speed. It's <unk>, it's not mostly I mean, we just have a deep appreciation for software you know, it's it's a combination of of architecture and software and and electronics design, you know chip design and that combination Melanie.

It's it's just world class and that's the reason why they're in 60% of the World Supercomputers, that's why they're in 100% of the A.I. Super computers.

And and the understanding of of large scale distributed computing was the second to none.

Now in the World and I, just talk about scale up.

And you're absolutely right now the question is why skill out and the reason for that as this this is the reason why they're doing so well.

The movement towards this aggregated micro servers.

Applications.

<unk> containers Microservices containers are distributed all over the data center.

And orchestrated.

So that so that the work load could be distributed across a very large hyperscale data center.

Architecture and you you probably know you know the three most important.

Application.

Yeah.

You know in my estimation in the world today.

No one would be tensorflow imply towards.

A number to would be spark.

And number three I mean kubernetes you know you could you could rank. It. However, you desire is and these three applications indicates a kubernetes.

It's a brand new type of application, where the application is broken up into small pieces and orchestrated across an entire data center.

And since it because it's broken up into small pieces and orchestrated across the entire data center.

The networking between the compute nodes becomes the bottleneck again.

And and that's the reason why they're doing so well.

By by increasing.

The the the network performance by off loading.

The communications off the C.P. is you you increase the whatever data center tremendously.

And so so it's it's the reason why they they had a record quarterlies quarters that we didn't want to be going 27% per year.

And their stock wish that there integration into the into the Hyperscale cloud Ah Ah Hyperscale club companies, they're low latency, they're incredibly low latency of their link mix mixed them really unique you been on you know whether it's either another incentive down in both.

And so so there there there it's a really fantastic stuck and then lastly lastly.

<unk>.

Yeah, we we would like to integrate we would like to innovate.

In this world, where where the world is is moving away from.

Just a C.P.U. as a computer node the new computing unit.

Software developer is riding a piece of software that runs on the imparted so.

In the future going going forward the computing the fundamental computing unit is an entire data center, it's so incredible.

Just utterly incredible you run an application one human could run an application and they were literally <unk>.

Activate an entire data center.

Yeah and.

World, we would like to be able to innovate from end to end from networking storage.

Security everything has to be secure in the future. So that we could reduce the reduce the attacks surface down to practically nothing.

And so networking storage security are all completely off loaded on incredibly low latency all incredibly high performance.

And all the way to compute all the way through the switch.

And then and then the second thing is we like to be able to innovate across the entire stuck you know then videos. It's just you know supremely upset that on software stacks and and the reason for that is because software creates market you can't you can't create new markets like we're talking about whether it's computational healthcare.

Or a thomas bribing or robotics.

Or conversational A.I. or Recommender systems, or you know edge.

All of that require software steps it takes software to create markets.

And so so our obsession about software and creating creating open class ones for the ecosystem and all of US all a bunch of over partners <unk> perfectly into that they they are they pioneered the open networking stuff.

And they pioneered in a lot of way sophomore defined data centers and so we're super Super excited about about the about the team and and and now we have the ability to innovate a in a data center scale world from end to end and then from top to bottom the entire stuff.

Yep.

Jensen.

They think line.

So next question comes from William Stein Listen Trust. Please go ahead.

Great. Thank you for kicking my question Jensen I'd like to focus on something you said I think it was in one of your earlier responses you said something about the very significant part of data centers are now now accelerated with G.P.U.'s I I'm.

Sort of curious how to interpret that if we think about sort of the the.

Evolution of computer architecture going from almost entirely.

Let's say racks and racks of.

P. P.U.'s too you know some future day, where we have you know many more.

<unk> and maybe a much smaller number of.

Of C.P. is relative to those.

Maybe you can talk to us about where we are in terms of that architectural shift and where you think it goes sort of longer term, where where we are the position of that.

Yeah I appreciate the question and this this you know for for for for.

<unk> computer.

Architecture, you know geeks and people who follow history, you know you know well that that in the entire history of time.

There are only only to.

Computing architectures that has.

As it has made it so far which is one of them as X. 86, the other ones aren't in any reasonable way.

And and if you get a on computer you get a good mix 86 computer you can program it.

And and.

In fact, there's no such thing as any sovereign computing platform until we came along and and today, where the only computing, it's already computing platform, but you could really marginally utterance, where we're in every cloud we're in every computer company.

There were a in every country you know we're every single side.

Ah, what where and we we accelerate applications from from computer graphics to video games to scientific computing to work stations to machine learning to about it.

This this journey to 20 some on yours.

Better company took 20 somewhat yours and we've been we've been focused on on X. already computing since the beginning of our of our company.

And and we made a general purposes. So we made a general purpose.

You know really starting with with a endeavor costs C.G.C. for graphics, and then it became khuda.

Yeah, and we we we've been working on phonics already computed for quite a long time and I think at this point is a foregone conclusion that it's already computing has reached a tipping point as well beyond.

The number of developers this year, though that that support we supported was almost 2 million developers around the world and it's growing what appears to be exponentially and and so I I think I think.

It's probably computing is now a well established.

Oh and video X. already computing as well establishes commonsense and and people who are designed data centres expect to put x. already computing in it. The question is how much.

How much it's already computed you do use and what part of the date and senior pipe when do you do it.

And and the big gigantic breakthrough of course, now we know well now.

And then video and videos recognized as one of the three pillars that the ignited the modern A.I. the big Bang of modern.

And the other people are of course is deep learning algorithm and the opponents of data.

And so the three d. three ingredients together and it was they you know people using video it's already computing largely for training, but over time, we expand training to have a lot more models and and and as I mentioned earlier.

The single most important model of machine learning today is the Recommender system.

It's the most important model because because it's the only way to you and I could use the internet and then he reasonable way is the only way that you and I could use a shopping website or or a video video web video Apple or music or book or you know news or anything.

So they just do it as the engine of the Internet from the consumers perspective on the company perspective, and injuring of commerce without Recommender system, There's no way they could possibly make money.

And so so they're accuracy in predicting user preferences is <unk> everything they do you just go up and down the list of every company.

<unk> engine is gigantic.

Just just a gigantic engine.

And from the data processing part of it which is the reason why we went and spent three years I'm spark and rapids, which made spark possible and all the work and we did on M.D. Logan.

Stuff was was really focused on big due to analytics. The second is all the all of the training with the deep learning models and then influence so the the number of applications.

The footprint celebrated computing has grown tremendously and it's important says grown tremendously.

Cause of the applications are the most important applications of these companies.

And so so I think I think.

You know I I think I think when I mentioned, when I said that that acceleration was just don't growing it is.

But the the major workloads the most important workloads up the world's most important companies are now are now alright, probably required celebration and so I I'm looking forward to it really exciting ramp for <unk> for all the reasons and I just mentioned.

So next question comes from John protect with quite credit. Please. Please go ahead.

Yeah, guys think show me ask the question was just too quick ones Collette I hate that something as mundane as topics, but just given the full your guide there there's sort of a lot to unpack and you talked about some of it like raises.

I think you also probably have some code with plus or minus isn't that I think there's an extra week. This year as well and then of course, there's mellanoxs and how you're thinking about investing in that asset.

Yes, I'm just kinda curious when when we look at the full your God is there something structural going on or not that says you try to take advantage of all these opportunities or can be used that as sort of a guide post to how you're thinking about revenue for the back half of the here as well I don't understand that and then Jensen just quickly for you Kinda makes sense to me that that Cove. It is accelerating activity.

He in sort of H.P.C., and and Hyperscale and maybe even in certain verticals like health care, but but in the other verticals as the the sort of shelter in place kind of hurt engagement and can we actually come out of covered with some pent up demand in those vertical markets.

Hey thing chunk question, let's start from the first perspective on the overall topic for the year, we've guidance and on gap at approximately 4.1 billion for the year, yes that incorporates three full quarters of Mellanoxs.

Mellanoxs and its employees, we have about close to 3009 ox employees coming on board you are correct. We have a 53rd week and this corridor excuse me not this quarter this year and that is husband outline.

Q filings that you should expect that as well we've pulled forward a little bit are focal nice several months in order to take care of our employees and then lastly, though we are investing in our business you see some great opportunities you've seen some great results from our investment and there's more to do.

We are hiring and investing in those businesses. So there's nothing different structurally but just this onset of mellanoxs and are investing together I think will produce long term part results.

And as usual John you know that that we're investing into into.

Yeah.

The I.T. industries largest opportunities.

Cloud computing and they aren't.

And and then after after these two opportunities is edging on.

And so we're we're we're we're looking down the fairway with with with some pretty extraordinary opportunities, but as usual you know we're we're we're thoughtful about the the rate of investment and and well well managed and.

And it is leadership leadership team is are excellent managers and and.

You know you could you could count on us to continue that.

<unk> was what was that Johns question 'cause you just give me one <unk> I have a typical.

D a of engagement levels in verticals, just with shelter in place it has that <unk> Oh, yeah right right.

Yeah, right a few some of the industry's up in effect.

We we already mentioned automotive industry beyond the automotive industry is there's been ground to to a halt.

Manufacturing has largely stopped and and and you saw that in our guidance, we expect automotive to be down 40% quarter.

It's not going to remain that way, it's going to come back and and.

<unk> nobody knows nobody knows what level is going to come back to an along but it's going to come back and there's no question in my mind. The automotive industry, you know the hunker down right now, but they will absolutely invest in the future of autonomous vehicles they have to.

No no no you know be extinct.

It's not possible not to have autonomy.

Capability in the future of everything that moves not not so that it could just completely drive without you.

That's that's a nice benefit too, but mostly because of safety and comfort and and just the joy of what seems like the cars reading Your mind, you know in and of course, you're still responsible for driving it and.

But but it just seems to be coasting down the road, you know reading online and helping you.

And so I I think I think the future of autonomy vehicles is is is a certain there's a certain.

People recognize the the incredible economics that that the the pioneer Tesla is is enjoying and and and and the industry is going to go. After it you know the the feature the future car companies that are gonna be software decline companies like any technology companies.

And they would love to have a an economic that allows them to enjoy doing school days with the sweets.

And so they're going to go after it and so this is uncertain that this is going to come back and well I I every conference and come back and and let's see the energy sectors are are I've been impacted the retail sector has been impacted.

There's there's you know those aren't large industries for us, but nonetheless the impacted.

The the <unk>.

The the impact in some of the industry's is accelerating their their focus and and robotics like for example on the one hand a B.M.W.'s.

Has has obviously impacted in manufacturing, which is the reason why they're moving so rapidly towards robotics, they have to figure out a way to get robotics into the into their guy cruise the same thing with retail even if we <unk>.

A lot more robotic robotic support in retail you going to see a lot more robotic support in in warehouses in logistics and and so during this time when their when their market when the industry is is.

Disrupted impacted it allows the market leaders to really lean into investing into the future and so when they come back there'll be coming coming back stronger than ever.

Thank you.

You know next question comes some not Ramsey with Colin Please go ahead.

Thank you very much good afternoon.

Two different topic tents minimal first of all congrats on amperes back of a product <unk>.

Well, but [laughter] at first question is it might have been a little bit hard to talk when to deal with pending about this topic, but now that it's close maybe you could talk a little bit about opportunities to innovate on and customize the mellanoxs dock and the balance of having an industry standard and the second one is.

You three cancel copy text moved around at the same time, there's there's obviously stay at home gaming demand.

Think about gaming product launch logistics and and any comments on there would be really helpful. Thank you.

Yeah. Thanks, a lot now appreciate the question I'll go backwards because because it's it's it's it's kinda cool on the one hand on the one hand I do Miss I do Miss that we we can't engaged the developers face to face there's just so much fun.

T.C. and all that work in the hundreds of papers are presented I learned so much you talk to each time, and then frankly I I you know I really enjoyed the the analysts me. So we have and [laughter] and so there's there's there's all kinds of stuff that I Miss about about the physical G.T.C., but here's the amazing thing we were in almost 58000 attendee.

The the G.T.C. kitchen keynote.

Yeah, I I did it from from my Kitchen, just right behind me.

And dedication keynote has been viewed almost 1 million times.

And and the the videos incredible, though so I think our reaches is.

Could be could could be quite great.

And.

I'm not I'm not too you know, we've got amazing marketing team and and just.

Got great people, they they're gonna, they're going to find a way to to reach or reach or gamers and and whenever we launch something next you know the gamers are gonna be in our in our customers are gonna be or.

Oh and markets are gonna be real excited to see it so I'm very confident that we're gonna do just what do we.

That what was the question before I should've never do that would just be industry standard versus customization mellanoxs opportunity I see okay.

There's there's.

We were so closely with mellanoxs over the years and and on the day that that we announce to T.C. you could see the number of products that we have a working together as a product synergies or really incredible and and the <unk> products energies include a lot of software development that they went in and.

In a lot of architectural development that went in DGX comes with nine no not smoke as I mentioned, if you look at our data center we ship.

We ship before we should give you access to the customers we ship it to our own engineers and the reason for that as because every single product that our company has a a in it.

Some jarvis to to Metropolis tomorrow and to try to you know Clara to Isaac to write all about products has has a a you know and and <unk>, where x. already frameworks for for all of the A.I. industry.

And M.P. or comes with a brand new numerical format called cheap Tenser float 32.

T have 32, it's just a fence has to be in the medical format and the performance is incredible and and we have to get an integrated and with the industry standard frameworks and now Tensorflow comes standard with with tens or <unk> T.F. 32, and probably torch comes standard with two what do you have to redo it and so we need our own large scale.

<unk> and so the first customer we ship to with ourselves.

And then we started shipping as quickly as we could to all of the customers.

You saw that in our data center in our Super computer we have.

170 570.

Maybe the R. brand new Mellanoxs switches.

And almost 1500.

200, <unk> Mellanoxs mix.

<unk> 15 kilometers of cables.

Fiber optic cables.

And and and that that is one of the most powerful super computers in the world today and.

Based on based on enter and and and so we we have a great deal of work that we did their together, we announced our first edge computer.

Between us and no one nods in this new card, we called the G.S.A. 100 integrate M.P.R. and then integrates Mellanoxs is C.X. six D.X., which is designed for by G. tell kosan edge computing, so incredible security and and.

Has a single wouldn't trust and and Virtualized and you know and so basically we we the C.G.S.A. 100, when you put into a standard center X. 86 server turns out server into a cloud computer in a box you can type the internet capability of a cloud of a state of the yard cloud, we're just cloud native it's secure.

It's it has incredible processing is now completely hyperconverged inside one box. The technology that made you just say 100 is really quite remarkable and so you could see all the different different product synergies that we have in you know working together, we couldn't we could've done <unk>.

Barkley so original without the collaboration with phenomenon. They worked on just on this on this piece of networking software called U.C.S. Oh, we worked on Nicole together it made possible infrastructure for large scale distributed computing I mean, it's just the list goes on and on and on and so we we the.

The two teams have great chemistry, the coach is a great culture fit I love working with them and right out of the shoot you saw you saw all of the great products energies that are possible because of the combination.

Oh.

That is all the time, we have for questions I'll turn the call that the chances for closing remarks.

Oh, it's coming.

I'm just.

Thank you, we had a great and busy quarter with our announcements we highlighted several initiatives.

<unk>.

Computing is moving to data center scale.

We're competing a networking go hand in hand, the acquisition of Mellanox gives us deep expertise and scale to innovate from entering.

<unk>.

They are is the most powerful technology force of our time.

Oh and pure generation <unk> several breakthroughs it is the.

Largest ever generational leap 20 add some training and in France throughput.

First unified acceleration platform for did analytics machine learning deep learning training and then things.

And the first elastic accelerator that can be configured for skill of applications like training to scale out applications like inference.

Amperes bat, it's universal and into last but.

It's gonna Rearchitect, the modern data center.

Third.

<unk> opening Marge new markets with A.I. software application framework, such as Clara for healthcare drive for autonomy vehicles, Isaac robotic Jarvis for conversational A.I. Metropolis with H.I.O.T. aerial for by G., and Merlin, where the very important recommender systems.

And then finally, we have built up multiple engines accelerated computing growth.

R.T.S. computer graphics.

Artificial intelligence.

And data center scale computing from cloud to edge I.

I look forward to update you on our progress next quarter. Thanks, everybody.

<unk> Conference call you may now disconnect.

[music].