Insights:

• Micron Technology achieved record-breaking fiscal Q4 and full-year 2025 revenue driven by AI data center growth.
• Capital expenditures reached $13.8 billion for the full fiscal year 2025, reflecting strong investment in future capabilities.
• Adjusted free cash flow was $3.72 billion for fiscal 2025, supporting financial stability and shareholder returns.
• The company declared a quarterly dividend of $0.115 per share, emphasizing its commitment to investors.
• Micron expects $1.2 billion sequential revenue growth in Q1 2026 with gross margins exceeding 50%, highlighting strong market momentum.

Micron Technology, Inc. closed fiscal year 2025 with record-breaking financial performance, largely driven by strong demand for AI data center memory solutions. The company’s Q4 results reinforced its leadership position in technology and operational excellence, setting the stage for continued growth into fiscal 2026. With a portfolio of high-performance DRAM, NAND, and NOR products, Micron is strategically positioned to capitalize on the growing AI market as the only U.S.-based memory manufacturer.

During fiscal 2025, Micron made substantial capital investments totaling $13.8 billion, demonstrating its commitment to expanding its manufacturing and technological capabilities. The company generated an impressive adjusted free cash flow of $3.72 billion, which supports ongoing innovation and shareholder value initiatives, including a recently declared quarterly dividend of $0.115 per share.

Looking ahead, Micron forecasts $1.2 billion in sequential revenue growth for the first quarter of fiscal 2026, with gross margins expected to exceed 50%. This outlook reflects solid market demand and confidence in its competitive product portfolio. Micron continues to drive innovation that enables AI, compute-intensive applications, and the broader data economy, positioning itself for sustained long-term success.

For more information visit

Micron Technology, Inc.

Insights:

• NVIDIA CEO Jensen Huang highlighted the UK as a burgeoning AI ecosystem with 3,700 companies and 60,000 employees.
• The UK-US collaboration involves the largest-ever tech agreement focusing on AI for security, innovation, and economic growth.
• NVIDIA committed £2 billion in investments with UK-based venture capital firms to foster AI startups and scale technology impact.
• Deployment of 120,000 NVIDIA Blackwell GPUs will power new AI “factories,” including partnerships with Microsoft, OpenAI, and others.
• Projects include building a quantum-GPU supercenter and launching a robotics R&D hub to train the next generation of AI talent.

This week, the United Kingdom emerged as a global AI hub during NVIDIA CEO Jensen Huang’s visit, marked by a high-profile event in London featuring UK Prime Minister Keir Starmer and senior government officials. The gathering emphasized the UK’s unique “Goldilocks moment,” where top universities, researchers, startups, and venture capital converge to create fertile ground for AI innovation. Huang praised the UK’s talented workforce and its historical significance as the birthplace of the industrial revolution, framing AI as the next major industrial leap.

The collaboration between the UK and the US, anchored by NVIDIA’s £2 billion investment and the largest-ever deployment of 120,000 NVIDIA Blackwell GPUs, aims to bolster AI infrastructure and commercialize cutting-edge science. This investment partnership includes leading VC firms and spans several cities beyond London, underscoring a nationwide impact. Key initiatives also include developing quantum computing capabilities, launching dedicated robotics research hubs, and accelerating major AI projects in healthcare, climate science, and public services, powered by the UK’s most powerful AI supercomputer based at the University of Bristol.

For more information visit

NVIDIA, Accel, Air Street Capital, Balderton Capital, Hoxton Ventures, Phoenix Court, Microsoft, Nscale, OpenAI, CoreWeave, Oxford Quantum Circuits, techUK, QA, University of Bristol

Cadence, Arm

For more information visit

Insights

• Eliyan has successfully taped out its NuLink PHY using Samsung’s SF4X 4nm advanced manufacturing process.
• The interconnect supports up to 40Gbps bandwidth with industry-leading low power density and advanced packaging technology.
• NuLink PHY enables custom High Bandwidth Memory (HBM) base dies and supports ASIC disaggregation for chiplet-based designs.
• This technology addresses critical memory and IO bottlenecks in next-generation AI and GenAI systems.
• Collaboration with Samsung Foundry highlights the integration of advanced logic and memory technologies for data center applications.

Eliyan Corporation has announced a major breakthrough with the successful tape out of its NuLink PHY in a x64 UCIe Advanced Package Module, manufactured using Samsung Foundry’s cutting-edge SF4X 4nm process. This new chiplet interconnect technology boasts an impressive 40Gbps bandwidth at remarkably low power levels, achieving unprecedented power density through the use of 45-micron pitch micro bumps. Fully compliant with the UCIe standard, NuLink PHY extends high-performance die-to-die and die-to-memory connectivity in both standard and advanced packaging formats.

Designed to enable next-generation multi-die semiconductor architectures, Eliyan’s solution supports the creation of custom High Bandwidth Memory (HBM) base dies and is available as IP to facilitate ASIC disaggregation strategies. This addresses the critical memory and IO challenges faced by AI and GenAI subsystems, particularly in training and inference applications for large language models and other high-performance computing tasks. The collaboration with Samsung Electronics reinforces the strategic importance of integrating advanced process technology with innovative interconnect design to meet the growing demands of data centers worldwide.

For more information visit

Eliyan Corporation, Samsung Electronics, Samsung Foundry