Nvidia has a roadmap for trillion revenue in the future, and having this company supply software alongside chips could hurt your wallet.
This $1 billion revenue projection doesn’t have a timeline, but it’s ambitious given that revenue in the most recent fiscal year was just $26.9 billion. But the company’s GPU technology conference highlighted that the path is through software, with hardware enabling it.
GTC has focused heavily on AI and graphics software spanning its GPU, CPU, data processing, and auto chip offerings. Nvidia’s belief is that in the long run, software will generate more sites from subscriptions and upgrades than money from one-time hardware shipments.
Nvidia CFO Colette Kress broke down the $1 billion revenue opportunity with $100 billion from games, $300 billion from chips and systems, $150 billion from AI Enterprise software, $150 billion from Omniverse Enterprise software and $300 billion from the automotive industry.
Nvidia CEO Jensen Huang’s keynote outlined a comprehensive system architecture approach to solving the buy-vs-build conundrum, a problem many non-tech companies face in emerging markets like automotive and artificial intelligence.
Nvidia is betting its offering of software, hardware and everything else will pay off, with companies lacking the in-house expertise to build local systems.
Some automotive companies, which are turning their expertise from manufacturing internal combustion engines into technology, initially rely on Nvidia for the underlying technology of automotive cars.
Nvidia is providing automakers with the means to implement Netflix-style pay-as-you-go services to enable features like self-driving. Mercedes Benz and Jaguar Land Rover will put self-driving cars with Nvidia’s computers and software on the roads in 2024 and 2025 respectively.
“These services offer OEMs an exciting opportunity to transform their business model as we’ve seen Tesla do with its Autopilot software priced from under $5,000 to $12,000 per car today.” , said Ali Kani, vice president and general manager of Nvidia. automotive division, during an investor day, which took place alongside GTC.
Software accounts for a large majority of the $300 billion automotive opportunity, Nvidia chief financial officer Colette Kress said at the investor meeting.
“Our software content per vehicle can be in the thousands of dollars over the lifetime of the vehicle, versus hundreds of dollars for hardware. And second, software scales with the installed base of vehicles, not with annual production” , said Kress.
Nvidia’s automotive business comprises three components: the Drive software stack for autonomous driving, in-vehicle hardware, and data center infrastructure for training and simulation, all of which have been upgraded at GTC.
Major automotive chip players like Renesas and NXP are largely focused on components, but Nvidia’s comprehensive approach relieves automakers of the stress of building complete AI systems for cars. Companies like Intel and Qualcomm also offer autochips, but largely work with partners, such as with the PC and smartphone markets.
Nvidia is also making big bets on the Metaverse, a 3D graphical universe where users can play, interact and build. Nvidia sees opportunities in hardware and tools for creation and collaboration in the graphics world.
But the Metaverse opportunity could be costly for users of Nvidia’s mainstream GPUs.
“The graphics needed to deliver a cinematic VR experience in a massive, physically accurate multiplayer world will likely require three to four orders of magnitude more performance than our highest-end GPUs.” Jeff Fisher, senior vice president of games at Nvidia, at the investor conference.
It also kicked buzzwords about virtual economies into the metaverse, with NFTs, billions in virtual real estate, and cryptocurrency as the backbone of it all.
“The GPU offers more value than ever. According to our data, they are spending $300 more than they paid for the graphics card they replaced,” Fisher added.
There are 3 billion gamers worldwide, and the number is growing every year, and there is no difference in the revenue opportunities of hardware (GeForce hardware) and software (GeForce Now in the cloud), said CFO Kress.
“Pricing per user is similar and annualizes to over $100 per year,” Kress said.
Nvidia also expects to generate nearly $150 billion in subscription software through enterprise use of Omniverse, a catch-all brand for Nvidia’s metaverse hardware and software offerings.
“We also estimate a $150 billion software opportunity based on two opportunities. First, a per-seat software subscription for professional designers and creators, which we estimate at $45 million [users] and second, a per-robot software subscription for digital twins, based on more than 10 million factories and warehouses,” Kress said.
Nvidia has made it clear that the AI and Metaverse software offerings work best on the company’s GPUs, CPUs, and other chips. Nvidia is also creating delivery mechanisms so that code can be delivered to Nvidia’s chips which could be on premises, in data centers, cloud servers or supercomputers.
At GTC, Nvidia also announced Enterprise AI 2.0, which is “the operating system for AI,” Manuvir Das, vice president of enterprise computing at Nvidia, told the conference. investors.
Compared to its predecessor, the new software has expanded hardware support with the ability to run on either CPU or GPU, as opposed to just the predecessor’s GPUs. The software will also run on virtualized and bare metal servers running VMWare, Red Hat or other platforms in major public clouds, compared to only VMware support in its predecessor.
“For Nvidia AI Enterprise, we estimate the total available opportunity at $150 billion based on the installed base of enterprise servers and our software pricing per server,” said CFO Kress.
Nvidia wraps its comprehensive offerings in “AI factories” like the EOS supercomputer, which will be operational in a few months. EOS is meant to be a showcase of its hardware, which includes its processor and networking stack. Nvidia at GTC announced the Hopper H100 GPU, which will be in the EOS AI supercomputer.
Nvidia’s software assets are built on the closed-source CUDA framework, which Nvidia considers a gem and a starting point on which Nvidia’s GPUs are built. Nvidia’s one-stop-shop approach differs from Intel’s and AMD’s open approaches with OpenCL, Intel’s OpenAPI, and AMD’s ROCm, given the reliance on partners for system integration.
Nvidia has tightened its grip on CUDA even further, making further improvements to the chips that ensure code written on them runs fastest on its GPUs.
At GTC, the company announced more than 60 updates to its CUDA software libraries, including frameworks for quantum computing, 6G networks, robotics, cybersecurity, and drug discovery.
“With each new SDK, new science, new applications, and new industries can harness the power of Nvidia computing. These SDKs tackle the immense complexity at the intersection of computing algorithms and science,” said CEO Jensen Huang during a keynote address on Tuesday.
The closed hardware approach creates a dilemma for potential buyers – go with Nvidia software and hardware or take alternative approaches. This consideration has already played out with some cloud buyers, with Google opting for a “build” approach with its own AI chips, and Facebook creating its metaverse with the “buy” approach with Nvidia’s AMD GPUs and CPUs.
Nvidia has also created faster communication channels for third parties to send code to run on its GPUs. The company has opened up the NVLink-C2C die-to-die interconnect so that outside chips can connect to its GPUs, CPUs, data processing units and other chips. The new Grace CPU Superchip, also announced at GTC, has processors connected using the NVLink-C2C interconnect. ®