NVIDIA CEO Jensen Huang has significantly expanded his company's strategic partnership with Marvell Technology, announcing a $2 billion investment to deepen their collaboration at Computex 2026. Speaking alongside Marvell chief Matt Murphy, Huang described Marvell as a potential 'trillion-dollar company', with Marvell stocks soaring over 30% following the announcement. The partnership focuses on connectivity solutions for next-generation AI infrastructure, addressing the critical need for high-speed interconnects as AI workloads become increasingly distributed across data centers.
The AI infrastructure transformation has propelled the global semiconductor market to unprecedented heights, with global semiconductor sales reaching $791.7 billion in 2025, representing a 25.6% year-over-year increase. According to the Semiconductor Industry Association, this growth momentum has continued into 2026, with Q1 2026 sales surging to $298.5 billion, a 25% increase over Q4 2025, and March 2026 alone hitting $99.5 billion, a remarkable 79.2% year-over-year increase. The industry is projected to reach $1 trillion in 2026, reflecting strong demand for AI acceleration, high-bandwidth memory, advanced GPUs, and AI-native infrastructure. This rapid growth underscores how AI has shifted from being a software trend to becoming a compute-driven industrial transformation.
Major semiconductor companies are experiencing record-breaking growth driven by AI infrastructure demand. NVIDIA reported record fiscal year 2026 revenue of $215.9 billion, up 65% from the prior year, with its Data Center division alone generating $193.7 billion, accounting for over 90% of total revenue and reflecting a 75% year-over-year jump in Q4 alone. Similarly, AMD posted record 2025 revenue of $34.6 billion (a 34% year-over-year increase), with its Data Center segment reaching $16.6 billion (up 32% year-over-year), driven by surging demand for CPUs and AI GPUs. The semiconductor industry is investing heavily in AI development, with US semiconductor firms investing approximately $119.5 billion in combined R&D and capital expenditure during 2024, including nearly $70 billion in R&D alone. AI and computer applications now account for roughly 35% of global semiconductor demand, highlighting the growing dominance of AI-centric infrastructure investments.
NVIDIA CEO Jensen Huang has highlighted the growing importance of connectivity in powering next-generation AI infrastructure, emphasizing that 'Useful AI has arrived' and is driving unprecedented demand growth. Speaking at Computex 2026, Huang explained that agentic AI systems require disaggregated and distributed computing patterns that create significant connectivity challenges. He noted that when computing problems are disaggregated into parts and distributed across entire data centers, connectivity becomes essential, making Marvell's technology crucial for scaling AI infrastructure. Marvell's silicon solutions connect everything from server components inside racks to geographically distributed networks, helping scale AI clusters without sacrificing performance. The industry faces growing challenges with copper cables hitting physical limits due to severe signal degradation and extreme heat generation at terabit data speeds.
Enterprise AI adoption is experiencing explosive growth, with the number of large AI-led deals increasing by 61% year-over-year, according to Avasant's Generative AI Services 2025 Market Insights™. High-tech and telecom organizations account for the largest share of Gen AI adoption at 20%, highlighting strong demand for AI-native transformation and intelligent application development. The enterprise technology landscape is undergoing a structural shift as organizations scale AI across operations, products, and value chains. As enterprises deploy generative AI, autonomous systems, and real-time inference use cases, infrastructure strategies are moving beyond generalized computing and centralized architectures toward domain-specific, distributed, and resilient models. This transition is redefining how enterprises design, source, and operate their technology foundations, with organizations increasingly aligning AI infrastructure investments with long-term business needs rather than traditional hardware refresh cycles.
