• June 5, 2026

This week I appeared on CNBC Asia to discuss several developments in the semiconductor arena, starting with NVIDIA CEO Jensen Huang’s announcement of its new chip architecture intended to “reinvent the PC.” I also talked with the CNBC anchors about Google’s ambitions for its TPU chips, as well as overall market trends for semiconductor stocks. While it was an engaging 15-minute discussion, CNBC hasn’t yet published a video clip online, so read on for an AI-generated, human-edited overview and summary of the segment.

Overview

NVIDIA CEO Jensen Huang announced a new computer architecture designed to meet the growing compute demands of agentic AI across PCs, laptops, and workstations. Analyst Patrick Moorhead highlighted NVIDIA’s expanding full-stack strategy, which creates a long-term competitive threat to Apple and others. The discussion covered rumored U.S. trade restriction updates on offshore Chinese entities, which Moorhead dismissed as having minimal impact, though he noted Huawei’s rising domestic AI capabilities despite trade constraints. Furthermore, Google is expanding its TPU ecosystem through neocloud partnerships backed by Berkshire Hathaway among others, positioning it well alongside NVIDIA’s dominant GPU market as AI demand remains robust.

Outline

Nvidia Agentic AI and PC Market Expansion

• The CNBC anchor introduces a clip from NVIDIA CEO Jensen Huang announcing a new computer architecture designed to meet the rising compute demands of widespread agentic AI.
• Huang asserts that this reinvention of the computer is as significant as the transition to smartphones, marking the launch of a brand-new product family across desktops, laptops, and workstations.
• He expresses gratitude that the entire global PC industry has partnered with NVIDIA to embark on this new line of products.
• The CNBC anchor asks Patrick Moorhead if he was surprised by the aggressiveness of NVIDIA’s move into the CPU front and its rapid revenue visibility.
• Moorhead notes that while the move was an open industry secret, he is highly impressed by NVIDIA’s multi-year roadmap targeting a 250-million-unit PC market.

NVIDIA’s Full-Stack Strategy and Competitive Dynamics

• Moorhead explains that NVIDIA is effectively extending its CUDA platform to play everywhere from datacenters to the industrial edge, cars, robots, and PCs.
• The CNBC anchor observes that NVIDIA is offering a full-stack solution encompassing the GPU, CPU, networking, and software, directly challenging rivals like Intel and AMD.
• Moorhead identifies Apple as a major long-term target, noting that NVIDIA’s new platform directly challenges high-end MacBook Pros and M5 chips used by developers.
• He emphasizes that NVIDIA brings a decade-long CUDA ecosystem and agentic platforms that, when paired with Windows, create the capabilities the market needs.

AI Cycle Realities and Datacenter Revenue

• Moorhead shares his perspective that the AI cycle is only in its second stage of a ten-stage journey, meaning that its acceleration is just beginning.
• He reflects on how NVIDIA previously spoke of replacing CPUs with GPUs, but notes that agents now require CPUs to act as the central conductors.
• Moorhead highlights NVIDIA’s recent financial disclosures, noting that the company added $10 to $20 billion of previously unseen datacenter CPU revenue.

U.S.-China Trade Restrictions and Huawei’s Innovation

• CNBC anchor Amanda Drury asks Moorhead about the impact of potential U.S. administration moves to close loopholes regarding advanced AI chip licenses for offshore Chinese entities.
• Moorhead dismisses the concerns as much ado about nothing, clarifying that the U.S. regulations have not changed since 2023, and stating his belief that the rumors will cause minimal or even zero impact for NVIDIA and major OEMs.
• Drury raises a philosophical point about whether over-restricting technology backfires by forcing China to build out and rely on its own independent tech ecosystem.
• Moorhead agrees strongly, stating that keeping China reliant on U.S. technology is a better lever, and notes that Huawei recently thanked the U.S. for the restrictions, which the Chinese company says forced it to innovate.
• Moorhead rates Huawei as the standalone leader in China, explaining that its advanced networking capabilities allow it to string together thousands of chips to rival NVIDIA’s performance.

Google TPU Neoclouds and Market Valuations

• The CNBC anchor shifts the conversation to Alphabet, asking how shareholders should view Berkshire Hathaway’s investment and valuation of the company’s compute story.
• Moorhead states that Google investors should be very pleased because the Google TPU represents a sophisticated alternative to GPUs, enabling margins of at least 50%.
• He describes a new arrangement where Google partners with entities like Berkshire Hathaway, Blackstone, and Apollo to create TPU-powered neoclouds to compete with NVIDIA-backed alternatives.
• Drury presents massive growth statistics for the Nasdaq 100, technology ETFs, and semiconductor indices, asking if the current AI market feels frothy.
• Moorhead concludes that unlike the dot-com bubble, this era is supported by real downstream demand, though a breather for the market is still possible.

Companies Mentioned

• NVIDIA
• Apple
• Microsoft
• Intel
• AMD
• Huawei
• Sugan
• Alphabet/Google
• Anthropic
• OpenAI
• CoreWeave
• Nebius
• Broadcom
• MediaTek
• TSMC
• Berkshire Hathaway
• Blackstone
• Apollo
• Qualcomm
• Dell
• HP
• Super Micro

Keyword Summary

NVIDIA, agentic AI, CUDA platform, data center, PC, GPU, CPU, data center, US China trade, Huawei, Google TPU, neocloud, Berkshire Hathaway

Recent Posts