SI Agent News
ConfirmedChips

Nebius returns to Nasdaq as a restructured AI infrastructure firm

The company rebranded from Yandex after offloading Russian assets to become a specialized provider of GPU compute and AI cloud services.

$18 billionNebius market cap before 2022 trading halt

The short version

Why it matters

This story highlights a unique corporate transition where a major tech entity successfully pivoted to a new identity and business model to survive geopolitical sanctions. For readers, it demonstrates how the massive demand for AI compute is creating new opportunities for specialized infrastructure providers.

A unique corporate rebirth

Nebius recently began trading under the ticker NBIS on the Nasdaq exchange. While it appears to be a new startup, the company has actually been public for 13 years. It originally floated in 2011 as Yandex N.V., the Dutch holding company for the Russian internet giant Yandex.

The company's trajectory changed drastically in early 2022 following Russia's invasion of Ukraine. Nasdaq halted trading on Yandex N.V. shares due to sanctions on Russian-affiliated companies. The company eventually successfully appealed to resume trading after a 16-month restructuring process to offload all its Russian assets.

The remaining entity, which was primarily a collection of infrastructure outside of Russia, was renamed Nebius AI. This new iteration is led by Arkady Volozh, the former Yandex co-founder. Volozh was removed from a European sanctions list in March after publicly condemning the assault on Ukraine.

Specializing in AI compute

Nebius has pivoted its core business to focus on AI cloud infrastructure. The company sells GPUs, or graphical processing units, as-a-service. These chips provide the necessary compute power for companies to run complex algorithms and execute machine learning models.

The firm recently debuted a cloud platform designed to support the entire machine learning lifecycle. This includes various stages of development such as data processing, training, fine-tuning, and inference. By providing these resources, Nebius aims to serve the high demand for specialized hardware in the current tech landscape.

Building this type of infrastructure is extremely capital intensive. Volozh noted that accessing public markets is the most efficient way to secure the necessary funding for such large-scale projects. The company's goal is to provide a stable platform for startups and larger enterprises alike to build generative AI applications.

Expansion and competition

Nebius is currently moving to expand its physical presence into the United States. The company recently announced plans for a new GPU cluster in Kansas City, which is expected to go live in early 2025. It has also opened customer hubs in San Francisco and Dallas, with a third planned for New York by the end of the year.

To balance its growth, the company uses a hybrid approach to infrastructure. It owns a data center in Finland and is tripling its capacity to 75 megawatts. Because building new facilities takes years, the company also uses co-location facilities in places like Paris to reduce latency for its customers.

In the competitive landscape, Nebius faces rivals from both hyperscale giants and other specialized cloud startups. One such competitor is CoreWeave, which has raised significant capital this year. While CoreWeave expands into Europe, Nebius is focusing on establishing a stronger foothold in the American market.

The broader Nebius Group

While cloud infrastructure accounts for two-thirds of the company's revenue, the Nebius Group includes three other distinct businesses. These include Avride, an autonomous vehicle company based in Texas, and Toloka, a Netherlands-based firm that specializes in data labeling and quality control for large language models.

Avride was formed from the international division of Yandex's self-driving unit. The team of 250 people transitioned to the new brand after the war halted their original plans to launch autonomous taxis in Moscow. Avride recently secured a multiyear partnership with Uber to bring its delivery robots and cars to the Uber platform.

The group also owns TripleTen, an edtech platform offering coding bootcamps. While TripleTen is currently breaking even and is not a primary revenue driver, it fits into a broader strategy. Nebius is exploring a model where it provides a full stack of services, ranging from data centers and hardware to education for the tech sector.

Market outlook and stability

The market's reaction to Nebius's return to the Nasdaq has been somewhat tepid. The company's market cap has fluctuated between $3.5 billion and $4.75 billion since it resumed trading. This is a significant decrease from its peak valuation of $31 billion in late 2021, but the stock appears to be stabilizing.

Volozh stated that the company could not predict the exact price per share because the situation was unprecedented. However, he noted that the stock has stabilized above the cost of the assets. This suggests that investors believe the company has a viable path to building a sustainable business in the AI space.

Looking ahead, Nebius must navigate the high costs of fleet building and data center construction. Volozh indicated that while the company is currently financing Avride, that unit will eventually need to find new partners. The company's success will depend on its ability to scale its infrastructure while managing its diverse portfolio of subsidiary businesses.

Watch for infrastructure growth

Investors and tech observers should watch for the completion of the Kansas City GPU cluster in early 2025. This will be a key indicator of Nebius's ability to successfully enter the U.S. market and compete with established cloud providers.

Additionally, the development of Avride's partnership with Uber is a significant milestone. The rollout of sidewalk delivery robots and self-driving cars on the Uber platform will test the commercial viability of the autonomous vehicle branch of the company.

Finally, the company's ability to maintain its minority stake in ClickHouse while growing its other entities will be important. It serves as a model for how the company handles spin-offs and capital allocation as it matures from a restructured entity into a dedicated AI infrastructure leader.

Sources and links

Liked it? Say why

Add a line for other readers.

No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.

Listen to the AI Wire headlines

Listen to the AI Wire

Listen to our daily AI shows every morning: on your commute, in the shower, wherever you are

Listen to the daily shows

First reported by TechCrunch · Read the original. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.

More AI news