Seligman Ventures Doubles Its War Chest to $1 Billion as the AI Boom Brings Hardware Back Into Venture Favour

Seligman Ventures, the venture capital arm of technology investment manager Seligman Investments, has doubled its deployable capital to $1 billion, citing faster-than-expected opportunities to invest in the infrastructure behind artificial intelligence. The firm announced the expansion on 28 September 2026, just seven months after launching with $500 million.

The move is a clear signal of how the AI boom is changing venture capital. For much of the past two decades, many venture investors avoided semiconductors and hardware, regarding them as too capital-intensive and slow compared with software. The enormous demand for computing power created by AI has reversed that view, and specialist hardware investors are now expanding rapidly.

From $500 million to $1 billion in seven months

Seligman Ventures launched in February 2026. At the time, it expected to make six to eight investments in its first 12 months. It has instead completed more than 14, taking six board of director seats and three board observer positions.

That pace led the firm to double the capital available for investment. Umesh Padval, a managing partner, said the range of opportunities across the AI infrastructure stack had exceeded the team's initial expectations.

"The opportunities we are seeing across compute, networking, connectivity, power, thermal management have exceeded our expectations," he said.

Who is behind the fund

Seligman Investments is part of Columbia Threadneedle Investments, the asset management business of Ameriprise Financial. According to the announcement, the platform manages about $48 billion in public and private market technology and healthcare investments, giving the venture arm access to deep industry research and relationships built over decades of investing in listed technology companies.

The venture team is led by chief investment officer Paul Wick, a long-time manager of Seligman's technology strategies, alongside managing partners Umesh Padval and Ashish Kakran. Eddie Ackerman serves as chief financial officer and operating partner.

Padval is a veteran semiconductor executive and investor with decades of experience in Silicon Valley, and Kakran has built a career in deep-technology investing. Both are part of a large community of Indian-origin leaders who shape the global semiconductor industry, from chip design and manufacturing to the venture firms that finance new companies.

Where the money goes

Seligman Ventures focuses on six areas: AI infrastructure, modern data-centre hardware, cybersecurity, cloud infrastructure, physical AI and robotics, and AI model labs including developers of world models.

Its named portfolio companies illustrate that focus. They include SambaNova, which builds AI processors and systems for enterprise workloads; Lumilens; Eliyan, which develops chiplet interconnect technology that allows multiple pieces of silicon to communicate at high speed; Upscale; Velaura AI; Cognichip; EPIC Microsystems; and Exaforce.

The breadth of Padval's list, from compute to networking, connectivity, power and thermal management, reflects the reality of AI data centres. Faster chips alone are not enough. The industry needs better ways to move data between chips, to supply enormous amounts of electricity, and to remove the heat those chips generate. Each of those bottlenecks is creating opportunities for new companies.

Why hardware is back

Several forces are driving renewed venture interest in hardware.

The first is demand. Technology giants are spending hundreds of billions of dollars a year on AI data centres, and they are actively looking for suppliers that can improve performance or reduce costs at any point in the system.

The second is exits. Large chipmakers and cloud providers have been willing to pay high prices for start-ups with valuable technology or teams. Recent months have seen significant acquisitions in AI hardware and models, including AMD's $8.2 billion deal for World Labs, announced the same day as Seligman's expansion. The stock market has also rewarded AI infrastructure companies with high valuations, raising the prospect of initial public offerings.

The third is the maturity of tools. Chip design has become more modular, with chiplets allowing start-ups to focus on specific functions rather than building entire processors. Design automation tools, increasingly enhanced by AI, can reduce development time and cost.

The risks of a crowded trade

Not every hardware bet will succeed. Chip start-ups require large amounts of capital and years of development before generating meaningful revenue. They depend on access to scarce manufacturing capacity at leading foundries. And they face formidable incumbents, notably Nvidia, whose products and software ecosystem dominate AI computing.

There is also the question of whether today's spending levels are sustainable. Some analysts have warned of an AI infrastructure bubble, arguing that capacity is being built faster than profitable demand can absorb it. If spending slows, start-ups selling into data centres would be among the first to feel the impact.

Seligman's rapid deployment suggests it believes the build-out has a long way to run. Its connection to a large public-market technology franchise may give it an advantage in judging those cycles, since its colleagues track the capital spending plans of the largest listed technology companies closely.

What it means for India

The expansion of hardware-focused venture capital matters for India's semiconductor ambitions. The country has become one of the world's largest centres for chip design, with most major semiconductor companies operating large engineering teams in Bengaluru, Hyderabad and other cities. A growing number of Indian and Indian-origin founders are starting companies in chip design, AI hardware and data-centre technologies.

Specialist investors with deep industry expertise and large funds are exactly the kind of backers such companies need. As Indian deep-tech start-ups mature, firms like Seligman Ventures could become an important source of later-stage capital and connections to global customers.

The bigger picture

Seligman Ventures' decision to double its capital is a snapshot of where investors believe value is being created in the AI era. The spotlight often falls on chatbots and consumer applications, but the foundations of the industry are physical: chips, cables, power systems and cooling.

With $1 billion to deploy and a team steeped in semiconductor history, Seligman is betting that some of the most important companies of the next decade will be built in those foundations, and that hardware, long out of fashion in venture capital, is once again at the centre of technology investing.