Even the market may care more about when NVIDIA, a $695B company joins the Trillion-Dollar club, let’s imagine how NVIDIA accelerate the 4th Industrial Revolution as a leading arms dealer.
Grace does it again
In Q1 2021, AMD reached a 10+% market share of X86 chips in data centers, beating Intel in Intel’s own game. In parallel, a fright of new territory powering cloud computing has begun, and NVIDIA is charging forward with GPU & ARM chiplets.
Grace Hopper was a pioneer of computer programming who invented one of the first SW linkers. In 2021, NVIDIA introduced a new chip product line, named Grace to honor her. This year, NVIDIA double downs with NVLINK, a chip-to-chip linker.
- Grace Hopper Superchip, a CPU & GPU NVLINK connected chip to compute giant-scale AI & HPC.
- Grace CPU Superchip, 2 CPU NVLINK connected chip to power HPC & cloud computing.
Coincidentally, it’s a similar path with Apple’s M1 Ultra lean innovation. Sure, it’s difficult for Intel, the incumbent to disrupt its x86 cash cow for the imminent term. But soon, it’ll be too late for Intel to catch up as the new growth reach the tipping point.
Mass Metaverse commercialization
Snow Crash, a 1992 novel forges the term “Metaverse” from “meta” and “universe”, a connected virtual world where billions may “stay for all the things they can be”. But “Rome wasn’t built in a day”, neither Metaverse. This GTC may just shows another roadmap toward that.
Recently, Meta CEO, Mark Zuckerberg & MS CEO, Satya Nadella are making big bets on Metaverse. Which attracts a lot of hype, and the tipping point may be in reach this time. Even it may not be obvious, NVIDIA is creating tools, infrastructure & platforms paving a way to the mass Metaverse commercialization brick by brick.
NVIDIA envisions “i am ai” to compute a better future built by prosumers & AI connecting physical & virtual worlds. NVIDIA is supplying Tanegashima (gun) for 3 big “killer app” domains.
- AI-powered science, engineering & business perdition
- AI-planned actions by Robotic Platforms: DRIVE, Clara Holoscan, Isaac
- AI-augmented Metaverse & Robotic SW: Omniverse Digital Twins
They may also found many bricks for Metaverse as a bonus. In a math from Nvidia CFO Colette Kress, $1T opportunity = $300B chips & systems + $300B AI & Omniverse enterprise SW + $300B smarter cars + $100B gaming. No matter when each industry reach the number, the tech investment will spill over paving the way.
Setting the accuracy of the estimation aside, how may we make sense of the implications in the long run? Let’s take smarter cars as an example. To realize Self-driving, a service provider needs a “multibillion-dollar NN training supercluster and 10+ billion miles of vehicle data” as Elon Muck suggests. No one comes close to Tesla’s full-stack innovations. Nvidia is now providing the missing “Dojo” pieces. On the way of Self-Driving, more cars will digitalize the world & more miles are driven in the virtual world to validate changes. This is not only enhancing robotic SW at an unprecedented speed but also creating higher fidelity Metaverses at a scale only limited by the computation power.
Full Disclosure
The opinions stated here are my own, not those of my company. They are mostly extrapolations from public information. I don’t have insider knowledge of those companies, nor a whatever expert.
