US chip-making firm Nvidia is reportedly expanding its focus to humanoid robots next year amid intense competition from peers in the AI chip-making business.

Nvidia’s vice president of robotics, Deepu Talla told the Financial Times that the company is expecting to introduce its next version of compact computers for humanoid robots during the first six months of 2025.

Nvidia bets on the coming robotics boom 

Nvidia has played a significant role in driving the thriving AI industry with its chipmaking business. Its chips are in demand worldwide as the demand for AI systems continues to grow. However, stiff competition is also arising within the sector.

This has pushed Nvidia to also look elsewhere. The $3.3 trillion company is now banking on the robotics boom projected in the near future.

According to Caliber, humanoid robots are expected to become more present in industries, warehouses, automotive factories, and even in some homes in 2025, although it is yet uncertain how mainstream they will become.

Now, the launch of Nvidia’s next version of compact computers for humanoid robots, Tetson Thor, will be a launchpad for its plans to be a top platform for this projected boom in AI robots.

“The ChatGPT moment for physical AI and robotics is around the corner.”

~ Talla.

Talla also told the Financial Times he believes that the market is at a “tipping point.”

According to the Financial Times, this also comes as the chipmaking giant is expecting more rivalry from the likes of AMD. At the same time, competition from cloud computing titans like Google and Amazon cannot be overlooked.

Nvidia is now channeling funds towards the “physical AI” space to help the new robotics firms grow. For instance, the company joined Microsoft and OpenAI in a February funding round that valued humanoid robotics company Figure AI at $2.6 billion.

Robotics business is a small fraction of Nvidia’s earnings 

According to the Financial Times report, the robotics business represents a small share of Nvidia’s total revenue, although the chipmaking company did not give a breakdown of the figures from its robotics segment.

During its third-quarter performance update released on November 30, Nvidia reported that data center revenue made up around 88% of the total $35.1 billion, which was 17% above the previous quarter and 94% from a year ago.

According to the company, data center revenue alone was 17% ahead of the previous quarter and 112% from a year ago.

Commenting on the financial statements then, CEO and founder Jensen Huang also acknowledged businesses were making investments into industrial robotics.

“Industrial robotics investments are surging with breakthroughs in physical AI.”

~ Huang.

According to a Cnet article, multiple humanoid robots are competing to secure their place in a crowded space, and next year is shaping up to be the “banner year for the marching metallic mechanisms made in our image.”

Companies intensified AI robot efforts in 2024

During the past year, firms like Boston Dynamics revealed a new all-electric Atlas robot, while retiring its iconic hydraulic Atlas – now known as HD Atlas.

The new robot can move in “superhuman ways.” Other companies, including Figure, Agility Robotics, and 1X, are also forging new partnerships with firms like OpenAI and Nvidia, using their hardware and software to assist robots in talking and performing different tasks faster than if they had to start from scratch.

Companies like Elon Musk’s Tesla have notably chosen to use the data, hardware, and software it was developing for driver’s assistance in its cars and repackaging it in a new form factor with its Optimus robots.

Meanwhile, in other robotics news, Pymnts also recently wrote about an MIT research that developed an AI system that could enable warehouse robots to handle odd-shaped packages and navigate crowded spaces without putting humans at risk.

This development comes when retailers and logistics firms are grappling with rising pressure to automate amid surging e-commerce demand.

The research shows that while robots excel at repetitive tasks like moving pallets, MIT’s PRoC3S technology tackles the long-standing challenge of robots handling more complex warehouse jobs safely.

“In theory, PRoC3S could reduce a robot’s error rate by vetting its initial LLM-based assumptions against more specific and accurate understandings of the warehouse environment,” Erik Nieves, CEO and co-founder at Plus One Robotics, told Pymnts.

“Think about it like this: A warehouse robot operating solely on LLM guidance has been described how to complete a task,” continued Nieves. He added that the PRoC3S goes a step further “by placing a digital robot in a simulated environment of that task.”

According to Nieves, this basically shows the difference between classroom instruction and “a really good field trip.”

From Zero to Web3 Pro: Your 90-Day Career Launch Plan