Omdia raises 2026 semiconductor revenue forecast as AI demand tightens memory and packaging supply

Omdia raises 2026 semiconductor revenue forecast as AI demand tightens memory and packaging supply

The research firm now expects 2026 semiconductor revenue to grow 94.1% year over year, with DRAM and NAND leading gains as bottlenecks in HBM, advanced packaging and leading-edge nodes persist into 2027.

Fact Check
The WSJ article 'ADP Targets More Gains Ahead After Quarterly Growth' confirms all elements of the claim: ADP forecast 5%-6% revenue growth for fiscal 2027, implying approximately $23 billion to $23.2 billion, and this broadly aligns with Wall Street estimates (FactSet consensus of $23.19B). Web search results also corroborate ADP's fiscal 2027 outlook of 5%-6% revenue growth.
    Reference1
Summary

Omdia raised its 2026 semiconductor revenue forecast to 94.1% year-over-year growth, citing stronger-than-expected AI-driven demand for memory and continued supply constraints across high bandwidth memory, advanced packaging and leading-edge foundry nodes. The firm said memory ICs are now expected to make up more than 50% of total semiconductor revenue in 2026. The supply squeeze is most acute in HBM, which is more complex to produce than standard DRAM and is manufactured at scale by only SK Hynix, Samsung and Micron. AI accelerators from NVIDIA, AMD, Intel and Google are intensifying that demand, while advanced packaging has become another choke point, with TSMC's dedicated lines running at full utilization and equipment lead times slowing expansion. Leading-edge capacity is also under pressure, with TSMC's 2nm and 3nm nodes largely booked by NVIDIA, AMD, Broadcom and Apple. Omdia said computing and data storage will be the fastest-growing application market in 2026, rising by more than 150% year over year to approach $1 trillion, driven by data center servers, memory-intensive workloads and higher memory pricing. The firm said higher memory and packaging costs are spilling into smartphones, PCs, consumer electronics, automotive and industrial markets. It expects premium smartphones, wearables, game consoles and OLED TVs to post meaningful semiconductor revenue gains, helped by rising component costs and sustained demand.

Terms & Concepts
  • high bandwidth memory: A type of stacked memory designed to deliver very fast data transfer, widely used in AI accelerators and other high-performance chips.
  • advanced packaging: Chip assembly techniques such as 2.5D and 3D packaging that improve performance by placing components closer together in a single package.
  • leading-edge nodes: The most advanced semiconductor manufacturing processes, such as 2nm and 3nm, used for cutting-edge processors.