Nvidia is often synonymous with GPUs, high performance processors originally designed to accelerate graphics rendering, which have more recently become the core of Artificial Intelligence (AI) model training and inference. By leveraging its AI expertise, over the past several years, Nvidia has developed a unique portfolio of AI software offerings aimed at democratizing AI, placing the company in a leading position to benefit from adoption of AI workloads.
SPEAR’s Differentiated View:
While many investors view the Metaverse as a science fiction or futuristic concept, we view it as a necessity for enterprises. Over the past year, Nvidia introduced its Omniverse offering, the core of its metaverse strategy, targeting its ProViz customers and Enterprise AI, targeting the vertical industries within the Data Center segment. We don’t see any competing product that comes anywhere near in terms of breadth and functionality and therefore believe that this could create a “winner-takes-all” scenario for Nvidia in the markets that it participates in. While investors have been focused on near term questions e.g. bitcoin mining demand, supply chain, gaming cycles, we believe they are missing a meaningful opportunity that lies ahead in AI software.
It is important to note that the opportunity here is broader than just the Omniverse platform and expect AI to be embedded in every application, every data center, and every product. Using what we believe are conservative assumptions, we arrive at ~$7-10bn software revenues by 2026, growing at double digits from that base, which we value at ~$250bn. This could be conservative as a high growth software businesses should trade at a significant premium to NVIDIA’s current multiple due to higher margins and more predictable revenue profile. We expect to see a commensurate hardware opportunity which will support strong growth for Nvidia’s existing hardware business providing downside protection to the current share price.
Several companies have talked about the metaverse and they all come from different perspectives e.g. social, gaming; Nvidia is focused on design and engineering, and we expect the company to capture a meaningful part of that market.
AI is in early innings of adoption and the “why now?” comes from the ease of deployment of Nvidia’s software offerings combined with improved compute capabilities that accelerated computing provides today.
Nvidia is well positioned to win both the software side, with its comprehensive software offering, and the hardware side, with its industry leading DPUs and GPUs.
Nvidia’s Metaverse: a meaningful software opportunity
For NVIDIA and its customers, the Metaverse is a platform where companies/users can interact, collaborate, integrate their current software tools, and enhance them with AI capabilities.
Enterprise AI will become a prerequisite to compete. Nvidia is in the process of making AI a mainstream workload for enterprise customers. The company over the past 1+ year introduced several software offerings which have been gaining traction as they solve some off the main challenges in AI adoption which are scalability and ease of deployment.
The company’s product offering can be purchased in several different ways that fit both small customers that want to run few applications and large customers that want to deploy AI for the entire organization. Omniverse targets Nvidia’s Pro-Viz segment and Enterprise AI targets vertical industries within the Data Center segment, with few different package options including Base Command and Fleet Command, and several vertical applications (see below our software framework analysis).
Using what we believe are conservative assumptions we arrive at ~$7-10bn+ software opportunity in the next 5 years (includes $3bn from Omniverse, $4bn from Enterprise AI, and upside from Base Command, Fleet Command and the vertical applications) growing at double digits going forward.
Exponential growth in hardware
AI model sizes are growing at an exponential rate and AI workloads are only at the cusp of their potential. Increased number of parameters means exponentially higher number of GPUs and DPUs, which is the core competency of NVIDIA. Any model with more than 1.3 billion parameters cannot fit into a single GPU. Models have grown in size from <100million parameters 2-3 years ago to 100s of billions today. NVIDIA and Microsoft just announced a 530 billion parameter model (the Megatron-Turing Natural Language Generation (MT-NLP)), which is the most powerful and largest monolithic transformer language model to date.
These large new models are in part possible because of Nvidia’s hardware offerings such as the DGX Superpod. DGX has a BlueField-3 DPU which is a 400Gbps data center infrastructure processor that delivers the equivalent data center services of up to 300 CPU cores, and has 22billion transistors.
Software opportunity valuation framework
Below we provide a framework for evaluating the incremental opportunity from Omniverse and Enterprise AI. Nvidia has compelling offerings in Automotive AI and Cloud Gaming which we will cover in future reports.
Omniverse is a collaboration platform made to create virtual scenarios and can be used to simulate cities, factories, airports, robotics, engineering projects, etc. Over 500 companies are currently evaluating Omniverse. Current users include BMW using it to achieve 30% more efficient throughput in factories, Bentley using it for designing and real-time testing of infrastructure projects and several others.
The benefit of Omniverse is that it is a metaverse with open standards, cloud native, multi-GPU scalable, and connects a large ecosystem of applications and software such as Adobe, Autodesk, Epic Games, Pixar, Blender, Trimble, Wrnch, Foundry etc.
The offering is priced on a subscription basis with 3 components: Nucleus, Creator, and Viewer. Nucleus is priced on per CPU socket ($5k) and each individual Creator license is $1,800 per user. An enterprise license includes Nucleus + 5 Creator licenses + unlimited Viewer licenses. Nvidia believes that there are ~20mm creators that could be potential subscribers (engineers, designers etc), so assuming a single digit adoption rate by 2026 implies a multi-billion-dollar opportunity in the next 5-10 years.
|NVIDIA Omniverse customers||1.6||3|
|Per user subscription||$1,800||$1,800|
Source: SPEAR Invest, NVIDIA
Enterprise fee for Nucleus incremental to this analysis
Three important takeaways: 1. Omniverse opportunity represents multiples of Nvidia’s current ProViz business, 2. Given the early innings of adoption, Nvidia could realize double-digit topline growth for 15+ years, 3. Software revenues come at a high margin, and therefore significantly higher earnings multiple.
Enterprise AI is a software suite comprised of solvers and libraries that Nvidia has developed over the years aimed at putting accelerated computing in the hands of every company. The company’s partnership with VMware vSphere significantly increases scalability and ease of implementation as 85-90% of IT today runs on VMware. This suite includes the horizontal layers of Nvidia software stack and will be sold along with vSphere in a standard software license model. It is priced at $3,600 per CPU socket + 900/license for support and gives customer access to NVIDIA middleware layer (CUDA-X).
Application specific vertical layers i.e. “frameworks” are sold separately. They include Metropolis for smart cities, Clara for AI powered healthcare, Riva for interactive conversational AI, Merlin for AI recommenders, Maxine for video conferencing, Isaac for robotics intelligence, and Morpheus for cybersecurity. Early adopters of these frameworks include Pinterest, Spotify, GE Healthcare, T-Mobile etc.
In addition to purchasing Enterprise as a software suite, Nvidia introduced a full offering which includes both software and hardware (DGX Superpod), where Nvidia manages the infrastructure and customers pay on a monthly basis. The goal of this offering is to reduce the barriers to entry for smaller customers. Base Command (platform for training and developing AI models) and Fleet Command (product that enables companies to deploy and manage the AI models out to the edge) both part of NVIDIA’ AI LaunchPad hybrid-cloud partner program.
We have built a framework to estimate the market opportunity for Enterprise AI based on industry data and datapoints provided by Nvidia. The vertical market and the infrastructure-as-a-service offerings (Base Command and Fleet Command) are incremental to this analysis as they are harder to quantify, but we believe could offer an equally large revenue opportunity over time.
|Server units shipped annually||12|
|Sockets per server||2|
|Enterprise share (excluding hyperscale)||40%|
|NVIDIA’s addressable market||9.6|
|Sockets with NVDA AI||2.88|
|Revenue per socket*||$1,260|
|Revenue per year||$3,629|
For revenue per socket, we use $900 annual maintenance
+ $3,600 per CPU socket amortized over 10 years
Source: SPEAR Invest, Nvidia