Counterpoint Research analyst Neil Shah said the rise of Agentic AI is reshaping the semiconductor industry by increasing demand for CPUs, GPUs, memory, and AI infrastructure as enterprises build gigawatt-scale AI systems.

AMD Expands Its AI Infrastructure Position

Shah said on Thursday that Advanced Micro Devices, Inc. (NASDAQ:AMD) delivered blockbuster first-quarter 2026 results as revenue rose 38% year over year to $10.3 billion, driven by strong demand for CPUs and accelerators.

He noted that AMD’s data center revenue climbed 57% to $5.8 billion, accounting for 56% of total revenue.

He said AMD strengthened its position as the only credible full-stack merchant alternative to NVIDIA Corp. (NASDAQ:NVDA) by offering CPUs, GPUs, FPGAs, networking, packaging, ROCm software, and rack-scale solutions.

Shah added that AMD aims to narrow the gap with NVIDIA, whose data center revenue still stands at roughly 7 times AMD’s.

Shah also said Agentic AI is reviving CPU demand because AI workloads now require more balanced compute architectures.

He highlighted strong traction for 5th-Gen EPYC processors across cloud and edge deployments and said AMD expects CPU revenue to rise 70% year over year next quarter.

He added that AMD’s upcoming 6th-Gen Zen 6 EPYC chips, built on Taiwan Semiconductor Manufacturing Co. Ltd.’s (NYSE:TSM) 2nm process, could drive further growth.

NVIDIA Maintains Leadership As AI Inferencing Gains Momentum

Shah said announcements at NVIDIA GTC showed AI inferencing taking center stage, with growing attention on token economics and efficient compute architectures.

He described NVIDIA and AMD as a “two-horse race” in full-stack AI infrastructure while noting that NVIDIA still dominates the market.

Shah said the upcoming competition between AMD’s MI400 series and NVIDIA’s Rubin Ultra platform will become a key battleground next year.

He also said NVIDIA continues to benefit from the broader AI infrastructure boom as hyperscalers and enterprises expand investments in GPUs, memory and rack-scale systems.

Intel, TSMC And Other Ecosystem Players Shape The AI Supply Chain

Shah said Intel Corp.’s (NASDAQ:INTC) supply issues and inability to meet CPU demand helped AMD gain server CPU market share and potentially surpass Intel in x86 server CPUs.

He added that ARM-based server CPUs are emerging as another strong alternative, although x86 architectures still dominate the market.

He highlighted Taiwan Semiconductor as a critical enabler of next-generation AI chips, saying competition for its N2, N3, and CoWoS capacity will intensify this year.

Shah also pointed to Samsung Electronics Co., Ltd.’s (OTC:SSNLF) HBM4 collaboration with AMD as an important memory partnership for future AI accelerators.

Beyond chipmakers, Shah identified Meta Platforms, Inc. (NASDAQ:META), OpenAI, Anthropic, Amazon.com, Inc. (NASDAQ:AMZN) Amazon Web Services, Alphabet Inc. (NASDAQ:GOOGL) Google Cloud, Microsoft Corp. (NASDAQ:MSFT) Azure, and Tencent Holding Ltd. (OTC:TCEHY) as major AI infrastructure customers driving semiconductor demand.

He also cited Wistron Corp and Foxconn Industrial Internet as important downstream manufacturing partners supporting gigawatt-scale AI deployments.

Price Action: NVIDIA shares were up 0.47% at $208.80 and Advanced Micro Devices shares were down 0.88% at $417.70 during premarket trading on Thursday, according to Benzinga Pro data.

Photo via Shutterstock