NVIDIA (NVDA - Free Report) delivered another strong beat-and-raise quarter last week and the stock once again stretched for new highs above $215.
On the top line, Q3 revenues surged 31.5% year-over-year to $2.636 billion, and surpassed the Zacks consensus estimate of $2.364 billion by 11.5%.
On the bottom line, NVIDIA posted earnings of $1.33 per share on a non-GAAP basis, up over 60% year-over-year, and clobbering by 41.5% the Zacks consensus of 94 cents per share.
The robust earnings performance not only stemmed from significant revenue growth but also from gross margin and operating margin expansion.
Is NVDA Too Expensive at 50X?
But many investors wonder if the stock is still a buy as it now trades for 53 times trailing 12-month earnings.
As an NVDA investor, I have a bias that says "Yes, I would still be a buyer near $200."
And, after this most recent quarterly report, as I go over the growth profile in all its different business segments and customers, I find multiple reasons to reinforce this view.
In sum, as long as the leading provider of gaming processors, High Performance Computing in data centers, and emerging AI technologies keeps delivering solid double-digit sales and earnings growth, along with upside surprises each quarter, then paying 48 times the current projection for next year's profits isn't a bad proposition.
So, two important questions follow...
1) Will the market demand for HPC/AI solutions in data center, gaming, and automotive segments continue to grow at high double-digits?
2) Can NVIDIA maintain their dominant position, even as Intel (INTC - Free Report) forges a renewed alliance with arch enemy Advanced Micro Devices (AMD - Free Report) to gain a foothold in GPU processor markets?
To help you decide whether the current 50X multiple is too expensive for NVIDIA's growth, this article will go over the current and forecast leadership of NVIDIA and why some Wall Street investment banks believe the company is on its way to earning $8 per share in 2020, which would equate to a 31X multiple at high Street targets near $250.
Every Major Dude Will Tell You
The other day I was reviewing all the big announcements at the Beijing GPU Tech Conference in late September, in particular that the giants of China e-commerce Alibaba (BABA - Free Report) , Baidu (BIDU - Free Report) , and Tencent were all new cloud/datacenter customers. I came away with the insight that "everybody is their customer now."
Then on Monday November 13, from the SuperComputing Conference in Denver, NVIDIA issued a press release highlighting the exact thing...
NVIDIA Chosen by Every Major Computer Maker, Every Major Cloud
DENVER, Nov. 13, 2017 (GLOBE NEWSWIRE) -- NVIDIA (NVDA - Free Report) today announced the world’s most advanced data center GPU — the NVIDIA (NVDA - Free Report) ® Tesla® V100 GPU based on NVIDIA’s Volta architecture — is available through every major computer maker and chosen by every major cloud to deliver artificial intelligence and high performance computing.
Dell EMC, Hewlett Packard Enterprise, Huawei, IBM and Lenovo have all announced Volta-based offerings for their customers. Providers such as Alibaba Cloud, Amazon Web Services, Baidu Cloud, Microsoft Azure, Oracle Cloud and Tencent Cloud have also announced Volta-based cloud services.
Building on this breadth of offerings, NVIDIA (NVDA - Free Report) has introduced new software and tools on the NVIDIA GPU Cloud (NGC) container registry that make it easy for scientists to deploy NVIDIA’s accelerated computing platform for compute-intensive research.
(end of NVIDIA press release excerpt)
Volta is NVIDIA's latest advance in HPC architecture. It delivers over 100 Teraflops per second (TFLOPS) of deep learning performance, over a 5X increase compared to prior generation NVIDIA Pascal architecture.
It's important to understand that NVIDIA doesn't just make HPC parallel processing chips for gaming, datacenter, AI, and other applications. The GPU chips with thousands of "cores" come embedded in some version of an overall platform known as CUDA (Compute Unified Device Architecture) which allows developers to immediately begin programming in a common language (say C++) what they want the machine learning system to do.
So the NVIDIA GPU Cloud (NGC) container registry adds a new level of service and ease of use for corporations applying the technology. In short, NVIDIA has just outdone themselves again in winning over the developer and enterprise communities by creating an extended platform which simplifies software integrations.
Deep Learning is NVIDIA's Business
Stick with me on mapping who NVIDIA's customers are and, just as importantly, why they are. If you understand this end demand for the technology from the "who's who" of big tech companies, you will better appreciate the potential value of NVIDIA's growth segments and the coming explosion of AI applications that I am going to describe for you in the last two sections.
So, how does NVIDIA speak to both individual developers and massive enterprises? After they build a CUDA architecture, they make sure that data scientists, entrepreneurs, and organizations can use it.
In May, the company announced that their Deep Learning Institute was expanding the number and availability of programming courses at little or no cost with the goal of training 100,000 developers by the end of 2018.
And on October 25, NVIDIA announced immediate availability of the NVIDIA® GPU Cloud (NGC) container registry for AI developers worldwide. Here was the word from that PR...
In just a few steps, NGC helps developers get started with deep learning development through no-cost access to a comprehensive, easy-to-use, fully optimized deep learning software stack.
The cloud-based service is available immediately to users of the just-announced Amazon Elastic Compute Cloud (Amazon EC2) P3 instances featuring NVIDIA Tesla® V100 GPUs. NVIDIA plans to expand support to other cloud platforms soon.
After signing up for an NGC account, developers can download a containerized software stack that integrates and optimizes a wide range of deep learning frameworks, NVIDIA libraries and CUDA® runtime versions — which are kept up to date and run seamlessly in the cloud or on NVIDIA DGX™ systems.
“The NVIDIA GPU Cloud democratizes AI for a rapidly expanding global base of users,” said Jim McHugh, vice president and general manager of Enterprise Systems at NVIDIA. “NGC frees developers from the complexity of integration, allowing them to move quickly to create sophisticated neural networks that deliver the transformative powers of AI.”
(end of NVDA PR excerpt from Oct 25)
Monday at the SC17 Supercomputing Conference (runs all week in Denver), where dozens of computer makers and cloud service providers launched a wide range of Volta-based products and services, CEO and NVIDIA-founder Jensen Huang proudly described their new achievement...
“Volta is the world’s most powerful platform for AI and HPC, and will allow the world’s top minds in scientific research to push the limit on what’s possible in areas like drug discovery, alternative fuel sources and predicting natural disasters. With Volta now in data centers and clouds around the world, a new wave of innovation is underway that will have an incredible impact across society.”
With "every major" computer and cloud provider adopting NVIDIA technology, the runway appears clear for continued 30% sales growth over the next 1-3 years.
How Far Will 50% Segment Growth Fly?
Before I break down NVIDIA's segment growth, let me share a few important notes on terminology: "Inference" is the term used to describe the next level of machine learning after "training."
Inference is sometimes referred to as deep learning. After scientists have "trained" a system on data, then they can have it apply the data set and knowledge to learn new things. Think of the initial training as establishing "weights" for data (baselines) and inference as advanced pattern recognition.
Gaming is still NVIDIA's biggest end market, claiming nearly 55% of the company's projected $9.4 billion in sales this year and still growing at over 25% annually. And gaming is still the largest global entertainment industry at over $100 billion in annual sales.
In September, I read several investment bank research reports which indicated that NVIDIA could continue to grow their cloud/datacenter business at over 50% annually for the next 3 years, with sales projections ranging from $7 to $10 billion in 2020 just for this segment.
Q3 results delivered confirmation of that optimism with Volta adoption driving datacenter segment sales to $501 million, representing 20% sequential (quarter-over-quarter) and 109% annual (year-over-year) growth.
NVDA noted that there are now more than 1,200 companies using its Volta-based inference platform, and this is why many Wall Street analysts are so enthusiastic about the continued growth in the datacenter space, especially since the total addressable market (TAM) is expected to push over $30 billion in the next 3 to 5 years.
Automotive is currently a small sales contributor for NVIDIA, at just under 10% of total revenues. But the segment is growing at nearly 50% as NVIDIA forms key partnerships in Europe (with Bosch and PACCAR) and in China (with Baidu) to deliver autonomous driving technology.
Even cryptocurrency mining is a growth category for NVIDIA, although it may never contribute more than 10% of total sales.
The Biggest Market Yet for AI?
GPU parallel processing architectures will continue to sweep business away from old-line CPU serial processing architectures in the computing market. Since gaming, datacenter, automotive and crypto-mining are all parallel applications growing strong double digits, NVIDIA has a clear path to double revenues in the next 3 to 5 years.
But there are advanced computing markets on the horizon that many investors haven't thought much about yet, and probably some analysts haven't even incorporated these markets into their models.
To highlight one of them, I've been in a surreal state of shock since I read a recent Bank of America research report about Augmented Reality. In it, the BofA/ML analysts calculate that last year's total market revenues tied to AR were $5 billion.
In 2020, they are forecasting this market to grow to $160 billion. That's a 3200% increase over 4 years! That's a CAGR (compound annual growth rate) of 138%!
I don't know what they see, but I like it.
Disclosure: I own shares of NVDA, BABA, AMD, and BIDU for the Zacks TAZR Trader portfolio.
Today's Stocks from Zacks' Hottest Strategies
It's hard to believe, even for us at Zacks. But while the market gained +18.8% from 2016 - Q1 2017, our top stock-picking screens have returned +157.0%, +128.0%, +97.8%, +94.7%, and +90.2% respectively.
And this outperformance has not just been a recent phenomenon. Over the years it has been remarkably consistent. From 2000 - Q1 2017, the composite yearly average gain for these strategies has beaten the market more than 11X over. Maybe even more remarkable is the fact that we're willing to share their latest stocks with you without cost or obligation.
See Them Free>>