Founder of Projetech, Inc.
The global artificial intelligence (AI) market is expected to reach the trillion-dollar mark by 2030, and just as it has done with the global automotive industry, Tesla looks set to absorb a considerable amount of market share. This is all thanks to Dojo, the supercomputer set to drive the most sophisticated (and fastest) AI training machine to date.
What is Project Dojo, and why does it matter?
Necessity breeds innovation: Tesla’s million-plus fleet of vehicles generates huge amounts of data, and the self-driving systems behind them require vast sums of real-world data. The computational demands for training these neural nets are huge, and since Tesla didn’t want to be limited by the general-purpose graphics processing units (GPUs) available, it decided to build something better.
It would be difficult to understate the significance of Tesla’s decision to go in-house. Not only is the company now developing bespoke hardware tailor-made for its specific needs, but it also represents a bold statement of intent to the tech monopolies currently dominating the AI hardware market. There’s a new kid on the block, and the speed of its development is a sight to behold.
Dojo was unveiled at Tesla’s AI Day last year, where Elon Musk implied it has the potential to reach exascale: 1 quintillion (1018) floating-point operations per second (flops), or 1,000 petaflops. In supercomputing terms, it’s the milestone to reach for.
MORE FOR YOU
Let’s start crunching numbers.
Let’s put these numbers into perspective. The earliest supercomputer (the 1964 Control Data Corporation 6600) could handle up to 3 million flops. It was remarkable for the time, yes, but approximately 147 billion times slower than the fastest known supercomputer today, with its capacity of up to 442 petaflops.
In short, Dojo has the potential to be the fastest supercomputer in the world—by a considerable margin. From a design perspective, its specifications are pretty mind-blowing. But the innovation goes beyond sheer computational power. Dojo’s combination of high bandwidth and low latency also gives it an operational edge over the competition, not just in terms of scalability but because it performs 1.3 times better per watt.
Why could Dojo be a game changer?
Dojo will work with Tesla’s fleet learning to harness massive datasets for training workloads. Its unrivaled performance and full vertical integration could drastically improve the speed and efficacy of the training, development and machine learning behind its autopilot and self-driving systems, further improving the safety profile of its fleet. Elon Musk suggested that by the time Dojo is fully deployed, traffic fatalities due to human error could be reduced by as much as 90%.
As an example, say there was an accident several miles ahead, the car would instantly inform upcoming vehicles to slow down. And that’s because Dojo will enable them to communicate with one another directly, with zero intermediaries.
According to Accenture, AI can boost business productivity by 40%, there is an insatiable demand for neural network training, and the introduction of the Dojo D1 chip could unleash a new wave of innovation across multiple sectors while mounting a serious challenge to the likes of Google, Cerebras and Nvidia.
With roughly a 90% market share in the supercomputer accelerator industry, Nvidia’s pioneering GPU technology hasn’t just outperformed the competition so far; it’s left rivals in the dust. So while Tesla is making major inroads with Dojo, Nvidia has itself been busy pushing the boundaries of AI in autonomous vehicles, with its much anticipated Drive Orin hitting production lines later this year.
Orin is a next-generation system-on-a-chip (SoC) capable of performing more than 250 trillion operations per second. First unveiled in 2020, it’s the new, in-vehicle supercomputer set to power Nvidia’s very own Drive software. It will bring real-time sensor data integration and allow autonomous vehicles to perceive and move safely through their environment.
A competitive AI market will massively expedite innovation across the board. For example, it will be possible to address complex global privacy and identity issues and drive incredible advances in medical research, healthcare, supply chain logistics and beyond. When game-changing hardware and software combine, the result will be incredibly accurate ways of analyzing data and predicting future outcomes across a range of industries and applications. A robust, healthy and hyper-competitive market stands to benefit us in ways we can’t even imagine yet.
While Dojo’s technology is being rolled out for use in Tesla’s self-driving vehicles, it could eventually be a leader in a new layer of cloud computing infrastructure: neural net training as a service. As our world becomes increasingly data-driven, supercomputers are set to play a starring role and unlock real-world cases—and usher in the next big leap in AI and self-supervised machine learning. It’s only a matter of time before fully autonomous self-driving vehicles are the norm.
Don’t bet against Elon. Learn from him.
From online payment networks to space travel to self-driving cars and now supercomputers, Elon Musk raises the bar at every turn.
Musk is a master of leveraging innovation capital to win support for his ideas. Tesla’s runaway success also comes down to its wider “ecosystem strategy.” As a company, it doesn’t just identify potential bottlenecks relating to a product’s individual components; it tries to solve them.
In the case of electric cars, batteries are the number one bottleneck. By investing heavily in battery research and producing them at scale, Tesla is betting big on controlling the bottleneck, and thus the profit center, for the industry at large.
The company wasted no time in expanding its charging network across the United States, thereby achieving a first-mover advantage. As a result, Teslas are the only electric cars able to cover large distances. Tesla has developed a fascinating, multipronged strategy for fundamentally changing an industry. It employs unique elements at each level of the ecosystem: overturning the core product architecture, positioning itself in key bottleneck components, and resolving system-level limitations that hinder adoption.
At the same time, Tesla has applied an effective approach to building the innovation capital needed to win resources, then deliver on its overall vision. Any organization can learn from this well-crafted approach.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?