Leading chipmakers are expected to unveil next-generation processors focused on on-device AI, with most designs relying on TSMC’s advanced 3nm, 4nm and 5nm process technologies to improve efficiency and performance.
CES 2026 is set to underline the semiconductor industry’s deepening focus on artificial intelligence at the device level, as AMD, NVIDIA and Qualcomm prepare to showcase new processors for AI PCs and emerging compute workloads. With the event continuing under the “AI Everywhere” theme, attention is shifting from cloud-heavy AI models to more efficient, locally processed intelligence across consumer and enterprise devices.
AI-centric processors take centre stage
AMD is widely expected to introduce new additions to its Ryzen AI portfolio, including updated desktop and integrated graphics processors designed to support AI acceleration directly on PCs. These processors are aimed at balancing higher performance with lower power consumption, a growing priority as memory and component costs remain elevated.
NVIDIA, meanwhile, is likely to use CES to outline the production status of its Blackwell platform while offering a clearer roadmap for its next-generation Rubin architecture. The company has been expanding its focus beyond data centres to applications such as agentic AI, autonomous systems and humanoid robotics, areas that demand high-performance yet energy-efficient silicon.
Qualcomm is also tipped to make the commercial debut of its Snapdragon X2 series at the show. Targeting Windows-based AI PCs, the new chips are expected to strengthen Qualcomm’s push into premium computing segments by enabling advanced AI features to run directly on devices.
TSMC’s advanced nodes in sharp focus
A key takeaway from CES 2026 is expected to be the continued reliance on Taiwan Semiconductor Manufacturing Company’s advanced manufacturing processes. Industry sources indicate that most of the new processors from AMD, NVIDIA and Qualcomm are being fabricated using TSMC’s 3nm, 4nm and 5nm nodes, highlighting sustained demand for cutting-edge foundry capabilities.
AMD’s upcoming chips are expected to leverage chiplet-based designs and heterogeneous integration, while Qualcomm’s Snapdragon X2 is reportedly built on TSMC’s enhanced 3nm process. NVIDIA also remains a long-standing partner of the Taiwanese foundry.
Intel looks to regain momentum
Alongside its rivals, Intel is expected to use CES 2026 as a platform to reinforce its client computing strategy. The debut of its next-generation Core Ultra processors, codenamed Panther Lake, is seen as a crucial test of Intel’s 18A process, even as parts of the chip are manufactured by TSMC. Intel may also unveil refreshed Arrow Lake processors for desktop and laptop platforms.
With advanced-node demand rising, supply-chain sources suggest TSMC is accelerating capacity expansion at its Fab 18 facility in southern Taiwan, reinforcing its central role in enabling the next wave of AI-driven computing.
See What’s Next in Tech With the Fast Forward Newsletter
Tweets From @varindiamag
Nothing to see here - yet
When they Tweet, their Tweets will show up here.
