This week shook up the foundation of tech infrastructure. Memory chips, the quiet backbone of AI, cloud, and everyday devices, are suddenly scarce, forcing companies to scramble for supply. Meanwhile, AI and quantum-tech investments surged again, signaling that the boom is far from over.

Some players are acting fast, reshuffling supply chains and capitalizing on rising demand. Others may soon feel the pinch. If you care about where AI, chips, and tech growth go next, this week matters big time.

Top 25 Moves

1. NAND shortage bites as AI eats storage
Phison’s CEO warned that global NAND flash supply is in severe shortage, with key 1-terabit TLC chip prices doubling from about 4.80 dollars in July to 10.70 dollars in November and suppliers fully booked through 2026. SSDs and AI storage arrays will cost more and ship later, hitting data centers, PCs, and embedded systems at the same time. AI is no longer stressing only GPUs. It is now structurally reshaping storage markets and forcing long-term allocation deals.

2. Micron commits 9.6 billion dollars to HBM mega-plant in Japan
Micron will invest about 1.5 trillion yen, roughly 9.6 billion dollars, to build a new high-bandwidth memory facility in Hiroshima, backed by up to 500 billion yen in Japanese subsidies. The plant will deliver crucial HBM capacity for AI training and inference but not before 2028. The HBM crunch will persist for years. Nations are using massive incentives to anchor next-generation fabs.

3. Moore Threads surges 400 percent in blockbuster IPO
Moore Threads debuted on Shanghai’s STAR Market, opening around 650 yuan after pricing at 114.28 yuan and raising nearly 8 billion yuan. The pop comes despite ongoing losses and export-control constraints. The company becomes a flagship for China’s domestic GPU ambitions. Capital markets are rewarding any path to GPU sovereignty.

4. India injects 4,500 crore rupees into homegrown chip lab
India approved about 4,500 crore rupees to modernize the Semi-Conductor Laboratory in Mohali and increase output by up to 100x. The renewed SCL becomes a national tape-out hub and training ground for chip-design talent. India’s semiconductor push is accelerating. Governments now treat fabs as strategic infrastructure.

5. Arm and South Korea open chip-design training hub
Arm and Korea’s industry ministry will train roughly 1,400 advanced chip designers through a new education center. The program strengthens Korea’s fabless ecosystem and supports giants like Samsung and SK Hynix. It also deepens Arm’s presence in Asia. Talent density is becoming the new competitive barrier.

6. Black Forest Labs raises 300 million dollars for FLUX models
Black Forest Labs closed a 300 million dollar Series B at a reported 3.25 billion dollar valuation to expand its FLUX visual foundation models. Backers include Salesforce Ventures, a16z, Nvidia, Canva, and Figma. The funds support compute scaling and a San Francisco expansion. Europe is asserting a real-world presence in the model race.

7. Quantum Systems becomes a triple-unicorn drone powerhouse
Quantum Systems raised an additional 180 million euros, bringing 2025 funding to 340 million euros and pushing valuation above 3 billion euros. Demand for its AI-enabled defense drones remains strong across NATO and Ukraine. The new capital supports air, land, and sea autonomy systems. Dual-use defense plus AI is becoming a stable investment category.

8. Model ML brings in 75 million dollars to automate banking workflows
Model ML raised a 75 million dollar Series A to build agents that generate pitchbooks, models, and diligence documents. If scaled, it shifts thousands of junior-banker hours from manual work to AI-supervised workflows. Banks are already testing it. Financial services remain a prime target for high-margin AI automation.

9. SF Compute raises 40 million dollars for its GPU marketplace
SF Compute closed a 40 million dollar Series A to grow its marketplace that rents spare GPU capacity. The approach boosts utilization and reduces infrastructure costs for mid-sized AI teams. GPU access becomes more liquid and market-driven. Compute trading is becoming its own industry layer.

10. Nevis launches with 35 million dollars for wealth-management automation
Nevis revealed a 35 million dollar Series A to automate up to 80 percent of operational tasks for financial advisors. Compliance, paperwork, and back-office workflows shift into continuous automated systems. Advisors can serve more clients without proportional staffing. Vertical AI is becoming the industry operating system.

11. Jeeva AI secures 9 million dollars to accelerate sales agents
Jeeva AI raised 9 million dollars to expand its agent that automates research, outreach, and CRM updates. The product already powers more than 35,000 users. It removes busywork rather than replacing reps. AI adoption in revenue teams is moving toward productivity-first strategies.

12. Marble Imaging raises 5.3 million euros for earth-observation satellites
Marble Imaging raised 5.3 million euros to build a high-resolution imaging constellation. Climate analysis, defense, agriculture, and insurance all benefit from sharper, more frequent data. Demand for upstream space intelligence continues to rise. Space-plus-AI is now a proven deep-tech investment theme.

13. Sokin secures 42.9 million euros for global fintech infrastructure
Sokin raised 42.9 million euros to expand payments and treasury services across enterprise and embedded-finance clients. Strong transaction volume justified the raise. Reliable financial pipes continue to attract large checks. Not all major rounds revolve around AI.

14. Milestone raises 10 million dollars to measure AI productivity
Milestone closed a 10 million dollar round to quantify how generative AI tools affect engineering output. Leaders get real data instead of guesses about AI’s value. Tool selection becomes performance-driven. A metrics layer is forming around AI adoption.

15. One-carbon Therapeutics raises 153 million SEK for oncology work
One-carbon Therapeutics raised about 153 million SEK, or roughly 16.2 million dollars, to advance oncology drug candidates. The platform uses computational biology to guide trial design. Biotech continues blending deep tech with traditional research. The sector gains resilience through data-driven methods.

16. Quantum Computing Inc secures 750 million dollars for photonic hardware
QCi completed a 750 million dollar private placement, bringing total liquidity to roughly 1.55 billion dollars. The funds scale thin-film lithium niobate photonic quantum systems and hybrid HPC integration. QCi can now pursue volume manufacturing. Quantum is entering billion-dollar capital territory.

17. DOE renews national quantum centers with 625 million dollars
The United States Department of Energy committed 625 million dollars to renew five National Quantum Information Science Research Centers. The program supports hardware, algorithms, and applied quantum projects. Long-horizon research receives stable backing. Public capital continues to anchor quantum progress.

18. Illinois deploys 250 thousand dollars for quantum-industry pilots
Illinois launched a Grand Challenges program awarding two 125 thousand dollar pilot projects that pair universities, quantum companies, and traditional industries. The structure ensures real-world collaboration. Quantum development moves closer to commercial use cases. Domains like manufacturing and microelectronics stand to benefit.

19. Booster Robotics adds funding after shipping 700 humanoids
Booster Robotics announced new capital after surpassing 700 humanoid robot shipments to more than 200 customers in over 30 countries. Major tech firms and universities are deploying the systems. Real-world deployments are accelerating past lab prototypes. Humanoid robotics is becoming a scaled export category.

20. SOSV gathers 200 robotics investors for a concentrated dealflow sprint
SOSV’s Robotics VC Founder Matchup brings 200 investors and 200 startups together for a full week of one-on-one meetings. The compressed format accelerates fundraising and pilot formation. Hardware founders gain faster capital access. Deal-making itself is becoming platform-driven.

21. Sierra adds SoftBank backing and expands into Japan
Sierra secured new investment from SoftBank’s Vision Fund 2 and announced Japan expansion with SiriusXM as an early flagship user. The move boosts credibility for enterprise-grade AI agents. SoftBank’s network opens strong distribution channels. Agentic systems are moving into real-world operations.

22. Baidu’s Kunlunxin prepares for Hong Kong IPO near 3 billion dollars
Kunlunxin completed funding above 2 billion yuan and is preparing for a Hong Kong listing around 2027 at about 21 billion yuan in valuation. Its P800 AI chips already support state-backed data centers. New M100 and M300 chips are planned for 2026 and 2027. China is formalizing its internal AI-chip champions.

23. Anthropic advances toward IPO with 10 billion dollars in projected revenue
Reports indicate Anthropic is preparing for a future IPO with expected revenue around 10 billion dollars and a private valuation above 300 billion dollars. Backing from Amazon, Google, Microsoft, and Nvidia ensures massive distribution and compute access. A listing of this size would redefine the frontier-model category. AI labs are becoming global-scale commercial platforms.

24. Bezos backed Project Prometheus launches with 6.2 billion dollars
Project Prometheus begins with about 6.2 billion dollars in funding and almost 100 staff to apply AI to aerospace, automotive engineering, and manufacturing. The company focuses on AI that learns from real-world physical trial and error. It targets sectors where digital agents cannot replace real physics. The next wave of AI will operate in factories, labs, and heavy industry.

25. Debate intensifies over how governments should allocate AI funds
Northern Gritstone’s CEO criticized the UK government for directing 5 million pounds to AI growth zones in Wales instead of concentrating resources in established clusters like Manchester, Leeds, Oxford, Cambridge, and London. The argument reflects a global dilemma. Policymakers must choose between high-density innovation hubs and broad regional support. Similar debates are emerging worldwide.

Full Insights

Insight 1: The Supply Shock That Will Shape AI for the Next Three Years

The past week made something very clear. AI is no longer limited by model size or GPU count. It is now constrained by the physical world. Memory and storage, usually quiet background components, have become loud bottlenecks that reshape entire industries. Micron’s decision to pour 1.5 trillion yen into a new HBM megafab in Hiroshima, combined with the global NAND shortage that has sent prices doubling since mid year, shows that the supply chain is entering a new phase of structural scarcity. India’s major investment in modernizing its national semiconductor lab confirms how global the race has become.

The Manufacturing Reality

Micron’s new HBM facility will take years to complete. The Japanese government may cover as much as 500 billion yen, but the plant will not meaningfully add capacity until 2028. Until then, the world will keep fighting for the same limited pool of high bandwidth memory.
At the same time, NAND shortages are deepening. Prices for key storage chips have more than doubled, and suppliers are sold out through 2026. This is not a seasonal blip. It is a structural shortage triggered by AI demand, long fab construction timelines, and consolidation among major storage companies.

The Escalating Cost of AI Infrastructure

Data centers rely on enormous amounts of DRAM, HBM, and NAND. When prices double, cloud expansion slows, hardware makers delay launches, and the total cost of ownership shifts upward. Every model that takes more tokens, more context, or more parameters consumes more memory.
Companies planning AI infrastructure for 2026 to 2028 will need to rethink budgets, contracts, and even architecture choices. Some will redesign systems to use less memory. Others will sign long term allocation deals that lock them into higher prices but guarantee they can ship products.

The Globalization of Fabrication Strategy

India’s investment into the Semi Conductor Laboratory is an important signal. Countries now view chip capacity as national infrastructure. Similar moves have happened in Japan, South Korea, the United States, and the European Union. The world is not moving toward a single global chip market. It is moving toward regional clusters that each want to secure their own supply, talent, and tooling.

Strategic Conclusion

The companies that win the next wave of AI growth will not be the ones with the best benchmarks. They will be the ones that secure memory, sign early supply contracts, redesign architectures for resilience, and diversify geographic fabrication partners. AI has moved from a software race to a hardware availability race. Leaders are already adjusting. Followers will feel the pain in 2026.