GL Signals - 2025 - A Year for AI Infrastructure/Verticalization, Robotics and Space
2025 is here! We’re sharing a summary of notable trends from the year 2024, drawn from a data review by our data and media partners and aligned with our investment thesis. These validate our market views in 2024 and tell us what to watch in 2025.
Crunchbase remarked: “This was the breakout year for funding to AI companies. Venture funding through mid-December reached $99.6 billion — up 80% year over year — Crunchbase data shows. Close to a third of that funding went to foundation model companies. The other two-thirds of funding went to sectors impacted by these new models. Separate from the hardware and data provisioning to manage and operate AI, leading sectors included autonomous driving, healthcare, robotics, professional services, and marketing and sales, Crunchbase data shows.”
Trend 1: Strong demands and investments in AI computing infrastructure, AI agents, and quantum computing; AI verticalization is the next wave.
From CBInsights: “AI’s cash grab - AI startups are grabbing 1 in 3 VC dollars. They’re also taking over the unicorn club — nearly half of companies getting $1B+ valuations right now are AI startups. And they’re getting to that mark faster (median of 2 years vs. 9 for non-AI unicorns) and with less headcount (203 vs. 414) … American startups now take 71 cents of every dollar of global AI equity funding. But, China’s big tech firms are backing startups that could threaten the leadership of US firms like OpenAI.”?
Also from CBInsights: “Big tech’s Capex heads toward $200B in 2025 - Tech giants are making massive investments in AI infrastructure like high-powered data centers. This spending is creating opportunities across the AI data center value chain.”?
Venture capital’s AI investment looks small compared to big tech’s AI infrastructure Capex spending (like what we wrote about Magnificent Seven one year ago). In 2024, many public companies released financials and guidance showing that the AI server and data center market is booming, while others (phone, PC, consumer, gaming, SaaS, automotive, etc.) are shrinking. Investment in AI infrastructure is an arms race among enterprises and countries, as it has the potential to affect national security and the global balance of power. A U.S. congressional commission recently proposed a “Manhattan Project-like” initiative for funding AI development to stay ahead of China.?
At the beginning of 2024, we wrote that “AI data centers are “picks-and-shovels”.... hardware is more a bottleneck than software now.” Now, the bottleneck is still driving the high growth of specific companies.
A semiconductor foundry market review shows: “While the AI revolution continues benefiting a few manufacturers, the rest are stuck in the industrial and automotive mud. … TSMC captures 90% of the operating profits of the Top 14 foundries.” TSMC has shown uniquely bright growth among manufacturers thanks to the advanced AI server technologies it provides.
Pitchbook also forecasted greater expansion of digital infrastructure in a recent report: “Digital infrastructure has been a huge topic this year (2024), with the proliferation of AI creating a tremendous need for compute, and thus, more AI-compatible datacenters. Investors have seen substantial returns as government initiatives coupled with the rapid digitalization of the global economy have given way to sizable growth. Digital infrastructure’s strong track record of performance and the confidence that it will continue have resulted in a cumulative $800.0 billion in commitments being made to private funds that engage in some degree of investment in the sector from 2014 to 2023.”?
According to Pitchbook’s report “H2 2024 VC Tech Survey: Investor Insights on AI, Dealmaking, and Fundraising,” investors are taking note of AI’s disruptive potential. Confidence in AI surged, with 47% of respondents expressing more bullish expectations compared with 34% of respondents in previous surveys. Significant shares of investors cited their key areas of interest as healthcare AI (39%), AI infrastructure (33%), and biotech (24%), as well as diverse “other” applications (30%) such as logistics, cybersecurity, and low-carbon solutions.?
From CBInsights: “AI agent landscape explodes - AI agents — LLM-powered bots that can independently execute workflows — have seen a massive increase in interest. Funding to AI agent startups more than doubled year-over-year in 2024. Over 40% of companies in the landscape have moved past the validating stage of commercial maturity and are actively growing commercial distribution.”?
Most people expect that AI will be applied to many verticals to make its next wave. The tendency for less regulation from the incoming Trump administration makes this even more promising.
On the computing tech frontier, quantum computing is also emerging. From CBInsights: “Quantum advances spur fresh interest in the sector - Equity funding to quantum computing companies has crossed the $1B mark 3 years in a row. A diverse cohort of startups is also racing to commercialize the still-nascent tech.”?
Trend 2: Power hunger of AI/HPC computing drives innovations in energy generation, cooling tech, and energy efficiency-boosting through connectivity, Si Photonics, etc.
Big tech companies are getting more involved in pushing clean energies forward for their AI data centers (although some may be greenwashing).?
All kinds of energy sources and new partnership models are emerging.
Nuclear power has been revived in 2024 thanks to AI data centers.
Sustainable computing benefits both the AI and clean energy industries.
Liquid cooling tech will see a high adoption and growth rate.
New methods for boosting energy efficiency are being developed, such as:
Deeper analyses can be found in our previous posts:
领英推荐
Most big tech companies (Tesla, Microsoft, Meta, Alphabet, and Apple) are trying to develop their own AI hardware (AI ASIC chips) to reduce their dependence on NVIDIA and improve energy efficiency. Recently, the public stock market has shifted attention to Broadcom over Nvidia. The market of AI ASIC chips is a new focus in 2025, and lots of startups are beginning to emerge!
There are many worries about how the incoming Trump administration will set back the climate tech industry. What is certain is that the energy transition has already begun. The deciding factor now is unit economics, not subsidies and policies. Canary Media recently developed charts to show the clean energy transition’s record growth.
Trend 3: 2024 has seen a strong rebound in robotics investments. Distribution is uneven thus far but may improve given that robotics is the next wave of AI.
Nvidia CEO Jensen Huang heralded the dawn of the robot era in 2024, claiming that robotics is the next wave of AI. To get ahead of this trend, the company launched Jetson Orin Nano Super, a powerful AI brain for robotics and edge, in December.?
AI being implemented into robotics, among other verticals, is a key development? to watch in 2025. A report from F-Prime Capital pointed out that “2024 has seen a strong rebound in robotics investments, but only for deals > $50M (mostly in defense and AV), early-stage deals (mostly focusing on vertical robotics) continued to decline.” Looking at the current political context, the incoming Trump administration's labor and immigration policies could create a tailwind for robotics in agriculture and manufacturing in 2025.
Trend 4: Space is the next frontier, with launching- and satellite-related value chains and applications seeing the biggest opportunities.
The number of space launches has quintupled in the last 5 years, while the cost to reach orbit has fallen by 8x since 2008. Cheaper access to space is sparking an investing rush and spawning startups across satellite tech, advanced propulsion systems, and more. Y Combinator publicly called for more space tech startups in September 2024.?
NFX posted: “We strongly believe that space is the next frontier. The space economy is a massive economic opportunity,? and launch companies will be the rent collectors of that economy.”
The proliferation of satellites is producing opportunities for related suppliers and satellite applications. Companies providing space infrastructure and support, including computing, communication, and transportation, may receive more funding or revenue. (More in this space to come in 2025)
Aerospace and defense sectors are also adjacent to space tech. Pitchbook reported that: “Deal activity in aerospace & defense is thriving, with the industry overcoming headwinds that have hampered the broader PE market. With an estimated 249 PE deals inked in the first three quarters, 2024 is on pace to set a record, according to PitchBook’s inaugural report covering the space.”
Trend 5: Hardware and semiconductors manufacturing reshoring for national security.
Just as AI’s strongest infrastructure opportunities and returns have so far lay in hardware and semiconductors, so too do energy transition, robotics, drone, satellite, and space launch technologies rely on hardware.. The role of software on its own is very limited, and the integration of hardware and software is crucial.?
Compared to software and AI, offshore manufacturing capabilities for hardware, advanced semiconductors, and deep tech are harder to replace. However, these capabilities have become a key national security issue, so onshoring them back to the US is in high demand. The escalating US-China cold war could also potentially create opportunities for suppliers in other countries, as US products may look to remove components manufactured in China.
What spaces are too crowded?
The report “H2 2024 VC Tech Survey: Investor Insights on AI, Dealmaking, and Fundraising” by Pitchbook found that “concerns persist regarding oversaturation in certain sectors - investors flagged chatbots, general-purpose AI tools, healthcare markets, and fintech markets as oversaturated spaces offering diminishing returns due to crowding and limited differentiation. Meanwhile, the commoditization of foundational models, such as LLMs, compounds these challenges.”?
This aligns with what CBInsights highlighted: “Fintech and digital health face sluggish VC environment - Funding and deals continue to sit at multi-year lows for the digital health and fintech sectors. The M&A environment is also largely stagnant … All 10 of the largest big pharma players globally have partnered with AI drug discovery startups since 2023. And 9 out of 10 are simultaneously developing in-house AI-powered discovery engines.”?
As a final note, the market may soon become crowded with AI agents. CBInsights pointed out:
Follow Global League here to get our signal updates, “GL Signals.”
Editor Bio
Jessie worked in the Taiwan semiconductor industry as an engineer and later as an R&D division manager, leading new semiconductor process/device development in her first career. Since immigrating to the US, she has been researching, consulting, and speaking for organizations on deep tech, AI digital transformation, and market insights across the US, Taiwan, and China. She has led professional community building for DoD ADL and IEEE from 2013 to 2018, and later for investors (Global League) collaborating on deal evaluation in AI and energy infrastructure, semiconductors, and national security. Jessie is committed to researching emerging tech and the market landscape and shares industry opinions for Digitimes in her spare time. Jessie now works for Enventure, a VC/PE firm investing in emerging tech startups in the US and Asia, and Global League (G-L.Ventures), an investor network.