Q3 2021 NVIDIA Corp Earnings Call
Good afternoon, My name is Jason and I will be our conference operator today at this time I would like to welcome everyone to Nvidias third quarter financial results Conference call. All lines have been placed on mute to prevent any background noise. After the speakers remarks, there will be a question and answer session. If you would like to ask a question. During this time simply per.
The star followed by the number one the on your telephone keypad. If you would like to withdraw your question press. The pound key. Thank you Simona Gen. Kempski you may begin your conference.
Thank you good afternoon, everyone and welcome to ingredient conference call for the third quarter fiscal 2021.
With me on the call today from Nvidia are Jen Hsun long, President and Chief Executive Officer, and collect Kress Executive Vice President and Chief Financial Officer.
I'd like to remind the better call is being webcast live on Nvidia of Investor Relations Web site. The webcast will be available for replay until the conference call to discuss our financial results for the fourth quarter of fiscal 2021.
The content of today's call is nvidias property, it cannot be reproduced or transcribed without prior written consent.
During this call we may make forward looking statements based on current expectations. These are subject to the number of significant risks and uncertainties and our actual results may differ materially for a discussion of factors that could affect our future financial results and business. Please refer to the disclosure in today's earnings release, our most recent forms 10-K and.
10-Q, and the reports that we may file on form 8-K, with the Securities and Exchange Commission. All our statements are made as of today November 18, 2020 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 you can find a reconciliation of these non-GAAP financial measures to GAAP financial measures in our CFO commentary, which is posted on our website with that let me turn the call over to collect.
Thank you Simona Q3 was another exceptional quarter with record revenue of $4.73 billion up 57% year on year up 22% sequentially and well above our outlook, our new Nvidia and Karen GPU architecture is ramping with actual.
Demand across our major market platform for Q.
Q3 was also in May and Mark quarter total for us and the industry of normal.
As we announced plans to acquire are from Softbank for $40 million. We are incredibly excited about the combined company opportunities and we are working through the regulatory approval process.
For today, we will focus our remarks on our quarterly performance.
Starting with gaming revenue was a record 2.27 billion of 37% year on year of 37% sequentially and ahead of our high expectations driving strong growth was our new Nvidia and pure architecture based G. for RTL 30 series of gaming Gpus.
The G. for RTL 30, 70, 38, and 39, the GPU offer up to two times the performance and two times for the power efficiency over the previous.
Turning based generation, our second generation, Nvidia Rts combined rate tracing and AI to deliver the greatest ever generational the performance for.
As announced on September 1st and ranging in price for $499 per $1499. The Gpus have generated amazing reviews, and overwhelming demand PC world call them Stocker Leanly powerful while newegg cited more traffic the.
The Black Friday, many of our retail and Etail partners sold out instantly. The RPX 30 series drove our biggest of launch for.
What we had anticipated strong demand exceeded even our bullish expectations given industry wide capacity constraints and long cycle times. It may take a few more months for product availability to catch up with demand.
In addition to the Nvidia Empiric GPU architecture, we announced powerful new tools for gamers as well as for tens of millions of lives streamers broadcasters E sports professional artists and creators Nvidia reflects is the new technology that improves reaction time in games reducing.
The system latency by up to 50% Nvidia reflex is being integrated into popular E sports games, such as App pens legends call of duty worsen fortnight and dollar rent.
In video broadcast is the universal plug in for video conferencing and live streaming of patients that enhances the quality of the microphones speakers and webcam with Nvidia AI effects, such as audio of noise removal virtual background effects and webcam audio free with debt remote workers.
Net of live streamers can turn any room into a broadcast studio.
Blockbuster games continue to adopt and videos are T X Ray tracing and AI technology at the games announced the fortnight, which has more than 350 million players worldwide is adding Nvidia RPX real time rate for spacing Nvidia CLS of AI Super.
The solution and in video of reflex, making the game more beautiful and even more responsive other major new titles featuring Archie ex this holiday season include watched of legend call of duty Black ops hold for a much anticipated cyber from 27 day selling gaming.
The laptop demand was also strong with double digit year on year of growth for the 11th quarter in a row Nvidia key for his laptop support the most demanding applications for creators and designers, while doubling as a powerful gaming rig by nine.
We also had a record of gaming console revenue on strong demand for the intent of switch.
And we continue to grow our cloud escalating in service GE Force now, which has doubled in the past seven months to reach over 5 million registered users chief.
Good for US now is unique as an open platform that connects to popular game stores, including the steam epic games and use of connect well now from gamers access to the title sales.
750 games are currently available on GFN. The most of any cloud gaming platform, including 75 free to play games with more of gains out of every sort of strength.
CFM supports many popular clients, including Pcs map and promo stay.
Stay tuned for more devices to come in the near future.
In addition, GFN reach continues to expand through our telco partners in a growing list of countries, including Japan, Korea, Taiwan, Russia, and Saudi Arabia.
We are also providing technology that enables the cloud in services to an expanding number of partners.
Following our earlier announcement with Tencent, Amazon and Facebook are beginning to offer cloud gaming services powered by Nvidia.
Moving to probe is.
Q3 revenue was 236 million down 27% year on year and up 16% sequentially ahead of our expectations.
Sequential growth was driven by strength in notebooks, which posted record revenue boosted by work from home initiatives and the shift to thin and light mobile workstation business, particularly offset by the decline in desktop workstation, which continued to be impacted by the pandemic.
And drove the year on year decline.
From an industry demand perspective, stronger verticals, including healthcare public sector higher education of research and financial services.
We continue to win new business and the number of areas in healthcare, we added the chronic for visual surgical applications and the results for medical imaging in technology and media and entertainment, we gained wins for design rendering and broadcast applications.
During the quarter, we announced the Ami birth, the world's first three the collaboration and simulation platform has entered open beta on the reverse.
Robert enables the tens of millions of designers architects and creators to collaborate real time on premises or remotely for using the virtual and physical world Ami Burns brings together in video breakthroughs in graphics simulation and AI. It will help enterprises address evolving.
Requirements as Workforces become increasingly distributed.
The initial market response from the transformative platform has been phenomenal over 400 individual creators of developers in the first industry have been evaluating on LIBOR and early adopters, including Ericsson BMW Foster and partners on the Lucas.
The pandemic is accelerating development of a AR VR and mixed reality technology, which will have a profound impact on the of how we work and play for example, our work with NASCAR to enable a variety of AR and VR services at the edge is revolutionary.
I mean, the racing experience for millions of fans across the globe.
With our industry, leading real time, Ray tracing graphics, AI and simulation of hardware and software stacks Nvidia is in the unique position to enable the future of blending the physical and virtual worlds.
Moving to automotive Q.
Q3 revenue was $125 million down 23% year on year and up 13% sequentially sequential growth was driven by a recovery in global automotive production volumes as well as continued growth in AI cockpit revenue for the year on year decline was due.
Due to the expected ramp down of legacy infotainment revenue.
In September Mercedes Benz debuted its redesign of EPS cost the Dan featuring an all new Nvidia powered Unbox AI.
Cockpit system with an augmented reality heads up display AI voice assistant and rich interactive graphics to enable every passenger in the vehicle to enjoy personalized intelligent features also.
Also in September the auto and beating electric cars for the China announced that it will develop its next generation of vehicles using the software defined Nvidia drive Asia ex foreign platform.
Orient delivers nearly seven times the performance and three times the energy efficiency of our previous generation associate making net uniquely capable to power. The next generation autonomous electric vehicles.
We have excellent traction with TV startups for.
Finally last week, Nvidia and Hyundai Motor Group announced the automakers entire lineup of Hyundai Kia and Genesis models will come standard with Nvidia drive in vehicle infotainment systems, starting in Twentytwenty two.
This feature rich software defined computing platform will allow vehicles to be perpetually upgraded with the latest AI cockpit features.
Now moving to the data center.
Revenue was a record of 1.9 billion up 162% year over year and up 8% sequentially driving growth was the strong ramp of our eight 100 based platforms continued growth with Mellanox and record chief for shipments for in France, Let me give you.
A little bit of color on each of our new Nvidia and tier architecture gained further adoption by cloud and Hyperscale customers and started ramping into vertical industry over the past weeks Amazon Web services, Oracle cloud infrastructure, and Alibaba cloud announced general.
Availability of the eight 100, following Google cloud platform and Microsoft Azure.
The 100 adoption by vertical industries drove strong growth.
We began shipments to server OEM partners, who brought whose broad enterprise channels reach a large number of end customers.
We also ramped the DGX 100 server and began shipping Nvidia DGX Super pause the first turnkey AI infrastructure. These range from 20 to 140, DGX 100 systems interconnected with Mellanox is HDR infiniband networking.
And enable customers to install incredibly powerful supercomputers in just a few weeks time.
In fact, we have announced plans to build an 80, no dts Super pod with 400, how to flow out of a outperformance called Cambridge margin, which will be in the UK fastest AI supercomputer the.
It will be used by and video researchers for collaborative research within the used case, AI and healthcare community across academia industry Amsterdam.
The joint the other systems in the videos complex of AI supercomputers powered by our R&D and autonomous vehicles conversational AI robotics graphics, HPC and other the domain.
This includes the lead now of the World fifth fastest supercomputer and fastest commercial supercomputer and the new Nvidia DGX Super pod, which ranks first on the.
On the Green 500 list of the world's most energy efficient supercomputers.
A great example of the tremendous opportunities for AI in healthcare is our new partnership with GSK for applying computational to the product and vaccine discovery process GSK for London based AI hub will utilize biomedical data AI method and advanced computing.
Not for me to unlock genetic and clinical data with increased precision and scale in.
In addition to this investment in video DGX 100 system GSK will have access to and videos, Cambridge run the Nvidia Clara discovery software and Nvidia scientists.
In Q3, the a 100 swept the industry standard ml per day benchmark for AI inference performance following our suite in the prior quarter's ml per benchmark for AI training, notably our performance led in AI inference actually extended compared.
With last year's benchmark for example in the rest net 50 test for image recognition are a 100 GPU beat CPQ only systems by 30 times this year versus six times last year. Additionally, a 100 outperform seek to use by up to two.
237 times in the newly added recommend data for test, which represents some of the most complex and widely used AI models on the internet.
Our winning performance in AI inference is trend for leading to continued strong revenue growth alongside the continued ramp of the eight 100 T for sales set a record as the Nvidia AI infringe the adoption is in total throttle.
We estimate that in videos installed GPU capacity for inference across the southern the largest public clouds now exceeds that of the aggregate CPQ capacity in the cloud Testament to the tremendous performance and Tcl advantage of our Gpus.
Hundreds of companies now operate AI enabled services on videos infringed platform, including the 100 or T for GPU, and our Triton and for and serving software for example, Tencent uses Nvidia AI inference to recommend video.
Music news and out supply.
Supporting billions of queries per day, Microsoft uses Nvidia AI in front of for Gram of correction in Microsoft office supporting half a trillion inquiries of year and American Express uses it for real time fraud detection.
We also gained tremendous traction in supercomputing, we announced that in video technology, including Amp here architecture, Gpus and HDR Infiniband networking will power five systems awarded by Euro HPC, a European initiative to build ex us scale.
Supercomputing the.
This includes Tri net comp a university consortium in Italy, and one of the world's most important supercomputing center, which will use and videos accelerated computing platform to build the world fastest AI supercomputer tightening the cost.
Supercomputer name minority of advances the age of ex the scale AI delivering 10 ex the plots of AI performance to enable AI and high performance computing converged applications use cases.
It has built with nearing 14000, Nvidia and pure architecture based Gpus and Mellanox HDR 200, gigabyte debt per second Infiniband networking.
And just the rig released Topfive hundred list of supercomputers show at Affinia, Gpus or networking powered nearly 70% and eight of the cash cost supercomputers on the list.
Mellanox had another record quarter with double digit sequential growth well ahead of our expectations contributing 13% of overall company revenue the of.
Syed reflected sales to a China OEM that will not recur in Q4 as a result, we expect the meaningful sequential revenue decline for Mellanox in Q4, though still growing 30% for last year.
Mellanox reached record revenue in both Infiniband and Ethernet driven by cloud enterprise and supercomputing customers strong demand for high performance Interconnects were mellanox as leader is being fueled by AI increasingly complex applications, which demand faster.
Monitor more scalable networks, but the data center becomes the new unit of computing in the age of AI Mellanox networking is foundational to modern scale long architectures.
At GTC in October we unveiled the blue feel to DP, you or data processing units, a new kind of processor, which offloads critical networking storage and security tiles from the CPQ a single Bluefield too deep you can deliver the same data center services that could soon.
Consume up to 125 CP Nucor's this phrase of valuable CPU cores to run a wide range of other enterprise applications. In addition, it enables zero Trust security features to prevent data breaches and cyber attacks and accelerate overall for.
For many of.
The M. were announced that it will offload accelerate and isolate is industry, leading the FXI hypervisor within videos the fuel to GPU boosting the sphere and data center performance and efficiency.
We also unveiled our three year GPU roadmap unified Mellanox is leading network capabilities with Nvidia GPU and the new Nvidia Udoka or data center on the chip architecture.
Software development kit for building GPU accelerated applications.
We believe that overtime the to use will shift of millions of servers unlocking a 10 billion total addressable market will appeal to the sampling now with major hyperscale customers and will be integrated into the enterprise server offerings of major Oems.
This was our busy period for product launches earlier. This week at Supercomputing 20, we announced the new double capacity, a 180, gigabyte Gpus and Gtx systems for organizations to build train and deploy massive AI models. We also.
Announced the new DGX station, a 100, a powerful workgroup server with for and one hundredg to use any massive 320 gigabyte GPU memory for data scientists and AI researchers working and offices research facilities labs or at home all of these.
Moving to the Nvidia Amp here architecture family of products will be available early next year.
As the 20, we also announced the next generation Nvidia Mellanox for 100 gigabit per second Infiniband architecture, getting AI developers and scientific researchers the fastest available networking performance the cells data throughput and as new any network computing engine.
To provide additional acceleration solutions based on this new architecture are expected to sample in the second quarter of calendar Twentytwenty one.
Moving to the rest of the PL Q3, GAAP gross margin was 62.6% and non-GAAP gross margin was 65.5% GAAP gross margin declined year on year, primarily due to charges related to the Mellanox acquisition, partially offset by product mix.
The sequential increase was driven by the absence of non recurring inventory step up expense related to the Mellanox acquisition.
Non-GAAP gross margin increased by 140 basis points year on year, reflecting a shift in product mix with higher data center sales, including the contribution for Mellanox non-GAAP gross margin was down 50 basis points sequentially in line with our expectations driven by product mix.
Q3, GAAP operating expenses were 1.56 billion and non-GAAP operating expenses were 1.1 billion up 6% and 42% from a year ago, respectively cash.
Q3, GAAP EPS was 2.12 up 46% from a year earlier and non-GAAP EPS was 2.91 up 63% for many years ago.
Q3 cash flow from operations was 1.28 billion.
With that let me turn to the outlook for the fourth quarter of fiscal Twentytwenty volume as a reminder, Q for include of 14th week, which we expect to be incrementally.
In addition to revenue and operating expenses.
We expect gaming to be up sequentially in what is typically a seasonally down quarter as we continue to ramp of our new Archie ex 30 series products.
We expect data center to be down slightly versus Q3.
With that we expect com computing products to grow in the mid single digits sequentially more than offset by sequential decline in Mellanox. We expect continued sequential growth in auto and pro business, so not yet returning to the year on year growth.
And we expect a seasonal decline in OEM.
Revenue is expected to be 4.8 billion, plus or minus 2% GAAP and non-GAAP gross margins are expected to be 62.8% and 65.5%, respectively, plus or minus 50 basis points.
The GAAP and non-GAAP operating expenses are expected to be approximately 1.64 billion and 1.18 billion respectively.
GAAP and non-GAAP.
Other income and expenses are both expected to be an expense of approximately $55 million GAAP and non-GAAP tax rates are both expected to be 8% plus or minus 1%. Excluding discrete items capital expenditures are expected to be approximately 300 million to 325 million.
Further financial details are included in the CFO commentary and other information available on our IR website in closing, let me highlight upcoming events for the financial community will be virtually attending the credit Suisse Technology Conference on November Thirtyth Wells Fargo TMT Summit December for.
First and the you'd be EPS tea and tea conference on December 7th our earnings call to discuss the fourth quarter and full year results is scheduled for Wednesday February 20 for.
We will now open the call for questions. Operator would you please poll for questions.
Certainly the at this time as a reminder of you would like to ask the question. Please press star one of the number one on your telephone keypad.
The pause for just a moment to compile the given day roster.
Your first question comes from the line of the John Pitzer from Credit Suisse. Your line is open.
Sorry can you hear me.
Yes.
Hey, guys. The resolution of the solves all thank you for me up the question just who.
Colette going back to your commentary around Mellanox. It seems like you're guiding the January quarter to about 500 million, which means the core data center business is still growing nicely call. It six 7% sequentially I'm just kind of curious when you look of the core data center business I know there is not a direct correlation to the server business.
But we're clearly going through a cloud digestion in server and core vertical markets enterprise for servers are we when you look at your core data Center business do you feel of build that's having an impact and use of sort of the digestion lets you saw kind of the bleach.
The school for 20 years is that the upside for school 19 in the 20, what you're doing is still growing significantly year over year or how would you characterize the macro backdrop.
Sure Let me let me share.
Clarify for those also on the call. Yes, we expect our data center revenue in total to be down slightly quarter over quarter. The computing products in video computing products.
The expected to grow in the mid single digits quarter over quarter as we continue the in video AI adoption and particularly as a 100 continues to ramp our net working our mellanox networking is expected to decline meaningfully quarter over quarter of sales to that China. Okay.
I.
I will not recur in Q4, though we still expect of the results to be growth of 30% for more year over year.
For the timing of some of the business therefore shifted of.
From Q4 to Q3 for overall H two is quite strong so in the referring to overall digestion. The hyperscale business remains extremely strong we expect hyperscale true growth quarter over quarter in computing products as a 100 continue.
True round. The 100 continues to gain adoption not only across those hyperscale customers, but again, we're also receiving great momentum in infancy with the a 100 and the T. for I'll turn it over here to Jon to see if he has more of that he would like to.
Yes, Hello, colon cancer of the doing well.
I want to go down.
Our our includes initiative is it's really gaining great momentum.
The answer is one of the hardest.
The signs problems compound the gigantic neural network computational graphs into our target. The margin is really really has proven to be really willing on the models. The models of diverse revenue provision to language to speech.
And on the so many different types of models being created the not bought model sizes are doubling every couple of months.
The latency of patients are increasing all the time for latency is decreasing all the time.
And so the question on on the.
On on imports is really great the technology pressures more great.
Our leadership there was the result is really pulling out of pulling her work for some generation terms for our team.
Well the over the course of last couple of years developing income server. It's called quantum has been adopted all over the place we have several hundred.
The customers now you do and the AI to deploy the non services.
This is from the early interest and I think this is going to be our largest near term growth opportunity for what we are really firing on all cylinders. The between the 100 is ramping the cloud of.
The 100, beginning to ramp in the mall enterprise.
And all of our influence initiatives are really doing corp.
In terms of maybe a follow on there just on the vertical markets clearly pulled from home and coated this year kind of presented the headwind to new technology deployments on Prem I'm kind of curious if we expect sort of an enterprise recovery in general next year, how do you think that will translate.
The into your vertical market strategy and is there anything else above and beyond that you can do to help accelerate penetration of the idea that end market.
Yes, John that's most of the point I mean that the the.
It's very clear that the inability to go to work.
Is the slowing down the adoption of side of.
The of the technology in some of the verticals of.
Of course of course.
Worsening lack of adoption.
Certain verticals like for example.
While the in health care too.
To rapidly discovering the that seems of.
Early detection of outbreaks of robotic application.
So warehouses.
Digital retail.
Last mile delivery of we'll just see just it's just really really great enthusiasm of loans.
The other you in on robotics technology on some of the old.
Some of the more traditional industries.
The new capabilities, the new technologies of slow so to the question one of the areas. The bolt on really Super excited about is the work that we're doing in the remote work until making the possible for people to collaborate remotely and weve apart from coal on the versions and some early data.
The good feedback from the marketplace as the really great. So I got a lot more to reported you guys.
The upcoming months around the room on boosting so.
But anyways I think when the where the industry of recover.
We sort of.
Our fundamental purposes, the company is to solve.
The the greatest challenges the impact the industry.
One of the computers cash.
And the challenges are are on our.
Sort of some of the most of whom applications in the verticals the way that we address and they're not commodity applications the wrong.
Really impactful.
Moving applications of cyber talking so when the when the industry.
The industries recover.
You know things with the non causal designed in claims of the loans and ships will be designs of buildings will be the ground.
On the consumer losses on the leasing a lot of simulation models of robotics applications.
Once again as a reminder of who would like to ask a question. Please press Star then the number one on your telephone keypad.
To give as many analysts of as possible. We used just ask one question. If you would like a follow up. Please re queue. Your next question comes from the line of the should you use from Evercore. Your line is open.
Yes, good afternoon, and thank you for taking the question.
You talked about in your prepared remarks limited availability of capacity components.
You suggested perhaps a few months could catch up curious if the if you can speak to the visibility that you have for both gaming and data center in into your April quarter.
Yes.
Glad you want me to take that because when the yep yep.
Yes, absolutely.
The DJ first of all the.
We have a lot of visibility into the into the channels of known especially for gaming and and we know how many weeks of inventory is in what parts of the channel.
We've been we've been.
We've been doing down the the channel inventories the timing for some time.
Meanwhile, we also expect that the very very successful launch of.
AMPU and and zone.
Even with our our bullish manage the station and all of the all of the.
The one because of the bill which is one of the fastest man ever.
The demand is still overwhelmingly.
And then I guess in a lot of ways of kind of expected for the circumstances are all of its been a decade since we've.
And then to the new type of computer graphics.
Two years ago, we invented the Columbus shooter and it's set the the industry on the course of.
The type of images, the we see today, but thats very clear the.
The teacher is going the wrote something much much more beautiful and the we invented and getting the Rts to do that and just to capabilities. One based on the truthfully in the of the ones based on artificial intelligence image generation.
Combination of those two capabilities is creating in the U.S, but the people were pretty excited about the.
At this point its decline dimension of instant council the.
So when we when we took the 10 years to invented.
We launched the years ago.
And took of second generation due to to really.
The level of the quality and performance the will be missed the interest.
The only when we really expect.
And now now the of the World. The the demand is just the overwhelming and so so we're in the we're going to one and two the ramp sales to lose maybe one of our most of social ramps Robert.
And the gives gives our install base of some 200 million plus cheap for steamers the best reason to upgrade.
Over the decades and so so this is the business is going to be of very large termination of course is my guess and then with respect the data center.
The ramping into a 100 of anyone.
100, as our sales force.
Generation of Gpus, but does several things of the same time is the universal the position of the universal because it's able to do all of the applications that the.
We we in the past has to have multiple gpus to do those training well the does influence incredibly well.
It does high performance computing and SaaS data analytics.
So so the table the NPL the AMPU architecture.
Able to do all of this of the same time and so the the utilization per for Datacenters.
For the utility is really really fantastic.
And the reception has been great. So so were going to ramp into.
All of the world's cloud of the starting this this quarter were now and will be for the major cloud provider in the world, including all the bomb Alibaba or goal.
And then of course, the the Giants the Amazon selling the measuring the coupons.
And.
Moving to continue to ramp as of that and and then of course, we are starting to ramp into enterprise, which.
In my estimation long term will still be the largest growth opportunity for us and we'll turn the every industry.
In two of the life altering the every company into the rise in interest.
Augments the within the volume.
On the deal.
You know the iPhone momentum to all of the world's largest industries. So we're ramping into that.
We're seeing great consumer clients.
Your next question comes from along the of.
She was going from Bernstein Research your line is open.
Hi, guys. Thanks for taking my question.
You said the of the extra week was contributing incrementally the revenue and Opex can you give us some feeling for how much.
It's contributing to revenue and Opex in Q4 of them and does that impact the lease on the revenue side. The upper say between my gaming and data Center and then how should we think about of impacting seasonality in the Q1 is that after we close on.
Sure Let me let me try the fund Johnson time, yes, good incorporated that 14th week and for our guidance for both revenue and Opex, we will likely have incrementally positive impact on revenue. Although it is tough to quantify okay. Our outlook also reflects incremental.
Opex for Q4 in primarily to different areas in terms of compensation.
Depreciation.
And given that our employees are such of material power of our Opex. It will it can be close to one for.
14th of the quarter now when we look a little bit for.
Further we should think about the incremental positive in both gaming and other data center front.
From the extra week as other hopefully will be extra supply, but not likely as much as 114 for the quarter of revenue as the enterprise demand.
Essentially project based and gaming demand, though is tied to the number of games that gamers that might be shopping for the overall holiday. So again still very hard for us to determine at this time normally between the Q4 and Q liner there is seasonality in the gaming.
Seasonality downward.
But we'll just have to see as we are still supply constrained within the Q4 to see what that looks like from an Opex standpoint, we'll probably expect our opex to be relatively flattish as we move from Q4 to Q1.
Your next question comes from the line of the Korea from Bank of America. Your line is open.
Hi, Thanks for taking my question on the congratulations on the on the Skunk work.
So my question is on competition from internally design products by some of your larger customers Amazon and Google and other via head about competition from time to time and I wanted to get your perspective is this a manageable risk is the right way to think that the perhaps using more of your.
In the public cloud, but they are moving to a ton of products for Internet of workloads, just how should we think about this risk going forward. Thank you.
Great Thanks for that.
Most of the cloud most of the club vendors and for the.
For the I believe all of the club members are use the same infrastructures margin for the return of program instrument.
For have the ability to or or largely do them and.
Those who the competition the competition.
We want to be really good and the reason for that exists.
Yes, the just the just the the acceleration.
To make the very clear that acceleration is the by towards the true.
And the non influence the.
The vast majority of the worlds.
The training models ours on doubling the size of the couple of months.
And it's one of the reason is one of the men so great on the.
The second is the this includes the.
The vast majority of the world's incurrence of balance of use.
The long term.
Nothing nothing is better.
The whole world looking margin, but the best way forward is to do improve on accelerators and when that happens our accelerator is the most versatile. It is the highest performance we moved the fastest a range of innovation as the market is the fastest because were also the most dedicated toward what most of which would have the most.
Just keep in the world to it.
Our stock is the most advanced giving us the the greatest versatility and performance.
And so so we see we see.
Oh the sponsors of announcements you are now.
But the also our largest customers going on and.
As you as you know the we're we're ramping.
Quite nicely a glowing.
The quite nicely and Amazon and Microsoft of our normal Workover and others.
So.
So I think I think the the big take away as Rob.
In the the great opportunity for most of the if you look at the vast amount of workload.
On the our workload in the world. The vast majority of the today is doing some views and it's very clear now.
This is going to be moving some of the the workload.
For the does the tone of the world.
Uh huh.
This is going to be really the core competency for us in the near term.
In fact, we believe it's on the largest growth opportunity in the near term.
And one of the early innings.
Your next question comes from the line of Harlan sur from JP Morgan Your line is open.
Good afternoon. Thanks for taking my question question and great job on the quarterly execution on the non of networking connectivity business was up 80% year over year I think it was up about 13, 14% sequentially and I know the roads the upside in October from non China customer, but the two girls, 70% fear of here.
Last quarter.
And you're still expecting 30% year over year growth next quarter, if I remember correctly I think infiniband is about 40% of that business Ethernet cloud is about 60% Jen.
And what are the big driver of the space, especially since we're in the midst of the cloud spending digestion cycle and I just saw that the team a non stared next gen 400 gig Infiniband solution, which should drive another strong adoption cycle. What's your supercomputer customers. When does this upgrade cycle start to fire.
Yes.
Let's see.
Our our data set of business consists of of.
Supercomputing centers, which is small.
High performance computing, which is which is a much larger part of supercomputing much larger the supercomputer.
And then.
Hyperscale and enterprise, which which.
About 50 50.
[music].
The other piece of the then kind of building of the business in the business.
The it's already computing part is not very much associated with digestion and others is much more so.
Associated with workloads.
Our new product cycles.
The Tcl that we bring in non inference.
The type of models of the the cloud service providers or the coin whether it is.
When the the deploying new non models based on the loan.
And how much of the debt, we how much how much of those workloads that we've.
We completed the according to one accelerators in.
The Oh, and the and ready for deployment.
So those are the factors associated with moving somewhere in computing, it's really it's really about.
About the Alps, it's really about the workloads and for them.
Moving on.
On the other half of the networking part of our business.
Is more is more connected to the CP business because the much more broad based.
The momentum part of our our business the Caribbean.
By the this idea of new New Hyperscale data center architecture called the dis aggregation of software does aggregation month, not necessarily the hardwood and softwood just like with the.
Where were the type of topical kubernetes or the street Michael services that are deployed across the data since the one service one application isn't monolithic one on one computer anymore, it's distributed across multiple computers and multiple nodes. So that the hyperscale data centers can more easily scale up and scale.
The allowance.
According to the workloads and according to the demand on the business.
And so this does this aggregation has caused the net working between the compute nodes to be of on by the importance and because mellanox is the lowest latency highest performing tires boom with network the new.
Yep.
The the Tcl brands the of the data set of skills really fantastic.
And so when the building our data centers.
No one else is going to be much more.
The month, we're committed to that.
On the enterprise side of the depending on on new.
New CP cycles.
Could affect on the CP cycle with the delay a little bit the would affect the by quarter, but would the Poland by quarter of it would affect them by by cooling of the quarter. So those are kind of the dynamics of the it.
I think the the net net of the is that the term.
Hi, it's a foregone conclusion at this point.
The AI is.
Is is.
There's going to be the future of the way software is written.
AI is the most powerful technology for the bottom line and and the acceleration is the best path for them and so that's what drives our computing business on the networking business has everything to do with the way the architecture, maybe some of the cloud data centers.
Which which is the which is not an architect of will mark the services now and Thats what.
Some additional interest the there.
The our networking business.
And so what we're really well positioned in these two fundamental dynamics because because as we know the on is the future income continues the future total goes both of those dynamics are very simple favorable to us.
Your next question comes from the line of Timothy Arcuri from the U.S. Your line is open.
Thanks, a lot of I wanted to ask a question of that the cooling.
But as the different way.
If I look at the core business, excluding Mellanox the core data center business.
It was up about 6% sequentially the past two quarters and your guidance sort of implies up about that much again in January which is certainly good on.
On the compound digestion, but of course, you have in peer still on the ramping as well, which should be a good tailwind. So there seems to be some offsetting factors. So I guess I wonder if you feel like your core data center revenue is still being constrained right now by some market died jastrem and kind of how you sort of balance for handicapped the.
The factors that.
Our our other.
Growth is tourism.
In the near term.
There's more affected by the cycle time of manufacturing.
And some of the flexibility of supply.
We we are we're in we're in good shape.
Two.
All of our suppliers in the sort of informs the on guidance.
But but we would appreciate it.
On shorter cycle times, we would appreciate.
More more agile agile on.
Supply chains, but.
But you know the world the world of the World is constrained at the moment and so so we'll just have to make the best of the the.
But even in that condition.
The non condition, we've only got us moving more on the.
Guidance Tom.
On the momentum we expect to grow.
Your next question comes from the line of earn workers from Wells Fargo. Your line is open.
Yes. Thanks for taking the question also congratulations on the quarter I wanted to go back to kind of the Mellanox question I know prior to the acquisition Mellanox is growing maybe in the mid to high 20% range here the last two quarters, it's grown over 75%.
Yes, I guess the simple question is how do you think about the the growth rate for Mellanox going forward and on that topic. We've started the here you talk more about bluefield and data processing unit I think of your commentary you alluded to serve.
Server OEM design wins, incorporating the CP is what are you looking at or when should we think about the GPU business really starting to inflect will become the a material driver for the business. Thank you.
Okay.
[music].
Long term long term.
Every computer in the world will be built like of Davidson.
And thirdly nodes on the data center is going to the data center and so.
And the reason for that is because once you had sort of this be basically zero.
And today, the most of the data center zone protected as the core free.
But in the future if you would like.
Cloud computing to be to be the architecture for.
Everything and every data center is multi tenant for the data centers the secure the you're going at the secure every single mode.
And each one of those notes are going to go into the.
Software defined networking software defined storage and it's going to have.
Per application security.
And and.
And so the processing done that as the you will have the will need to offload.
Of the CP use is really quite significant in fact in fact, we.
We believe that somewhere between 20% to 40% of todays data centers cloud data centers.
The the capacity of the throughput the the competition the load is consumed running basically the infrastructure overhead.
And that's what that's what the DP use internally was designed to do we're going to we're going off flow down number one the number two we're going to make every single application secure and competence of computing zero of course computing and will become a reality.
So the the importance of pneumococcus quite tremendous and I believe the record the every single.
Server in the World War, how the Gpus ensemble.
Just just because we care so much of our security interest because because so much about throughput and tcl.
But it's really it's really the most cost effective way of getting the data center.
In the winter so.
So I expect the I spoke of.
Our GP business to be to be quite large and so.
So thats the reason why we put so much energy into the city.
It's a programmable data center on the chip in the not running into the some of the infrastructure on chip.
And it's the reason why of order would be somewhere.
On on the taking.
Taking the the operating system of the center the softwood upon the opposition to the being sort of putting on on the on Brookfield and so so this is this is a very important initiative for us Im quite excited about it and you can imagine.
Your next question comes from the line of almost Triple storm.
Sure the star of from BMO capital markets. Your line is open.
Yes, thank you very much.
Colette and I apologize if I missed the but for Mellanox do you expect it to the.
Get back to the to go.
Sales for the three on a sequential basis in the quarter.
And I'm, assuming that the the shortfall in the current quarters for pulling from from all the.
Okay.
So our impact to our Q4 guidance for Mellanox, Yes is impacted.
By a sale two of China OEM for Mellanox that will not recur in Q4 as we look for the into Q1 of April were going to take this of quarter at a time and provide thoughts and guidance for that once we turn the corner to the new the new for us.
Fiscal year.
At the highest level club I think the the it's.
For the safe to say the high speed networking is going to be one of the most important things in cloud data centers.
As we as we go for and.
The vast midyear with the world's data center is still built for the traditional hyper converged architecture.
Which is all moving over to Michael services based the savages software declines the so I would give the architectures. The Mad journey is is still in its early days and so so I fully expect.
Future cloud data centers, all future designs.
Going to be connected with high speed networking inside the call of east West traffic and.
All of the type of tool for secure and.
And so so imagine the imagine building firewalls into every single server and the managing every single day.
Every single transaction every single transmission inside the data center to behind the and fully encrypted and so putting amazing amount of competition is going after the installed into into future day becomes but thats. It thats it.
That's a that's an except the requirement now the ones.
I think our networking business momentum is in the early innings of the group.
Your final question today comes from the line of of William Stein from true of Securities. Your line is open the great. Thanks for taking my question for you.
One of some pieces of this.
Puzzle, but I'm, hoping maybe you can address correctly the.
Sort of.
SKU by SKU rollout of the appear we know that.
We didn't have a ton of skews the last quarter there were more in the quarter. The just announced now you're doing sort of the.
Refresh it sounds like with double the memory of the Inc.
100.
Is the t. for going to be refreshed and if so why.
When did that happen and other other either.
Either systems or chips that are still waiting for the ampera refresh the could potentially contribute to an extended cycle as we look at the next year.
Yes in terms of in terms of the total number of skews the we've ramped of.
And pure for we're probably somewhere along a third to a half of the of the skews. The at this point, maybe a little bit less.
Yes.
The way the you could be true you could you could reverse engineer. It is like the you know we know what our what our gaming.
Our gaming line of looks like for desktops.
And so traditionally we we try to have a new architecture in every single segment and we have non we've not on the loan gone below for the non yet and so.
So there is there is a very big part of the marketplace. The we're still in the process of of.
Addressing.
Then the second thing is on.
Laptops, none of those none of the New York and true.
The architecture has has launched the laptops and the moves workstations do something with the pumps and workstations and laptops workstations and among them none of those of the.
Gone on here most of our data center.
In our data center business for for.
For cloud you've seen some of the early versions of the.
The 100.
And then there's the zone.
Cloud computing for Grafix, there's continuing.
Cookers enterprise.
Edge enterprise applications of enterprise data analytics applications.
And soon so there is a there is a fair share a number of exciting new clients. We we still have the Columbus.
That concludes our culinary for it so I now turn the call back to the is can you kind of skew for closing remarks.
Actually that will be for Jen hsun.
My apologies.
Okay.
Thank you. Thank you for a month.
This this was a terrific quarter and.
Many of our firing on all cylinders and the the.
Our tiered says reinvent the graphics on the has made real time retracing the standard Mexican constant current.
The best ever reason to upgrade for hundreds of millions of Nvidia gamers.
Hi will talk were write software no humans can is the most powerful technology for spar timing impacting every industry Nvidia.
The media against sweat ml per training.
And.
Now influence as well extending our leadership in this important new way of doing the computing.
Many of you try to influence server.
The other apply point the.
I will speak a lot more about in the future.
And a lot more frequently because of its important to in our full stack optimize platform of Guinea rapid adoption to operate many of the world's most popular enhanced services opening a major growth opportunity data centers are the new unit of computing.
Some day, we believe the will be millions of autonomous data centers distributed although for the globe.
In video of Bluefield, GPU programmable data center on the chip and our rich software stack will help close AI data centers in factories warehouses, Fiveg base stations and even on wheels.
And with our pending acquisition of acquiring the company the builds the most of the world's most popular CPQ, we will create the computing companies for the age of the in line with computing extending from the cloud.
So the trillions of devices.
Thank you for joining us today I wish all of you a happy holidays increases the safe and I'll look forward to see you guys next time.
That concludes today's conference call you may now disconnect.
[music].