Nvidia CEO Jensen Huang delivered the keynote address on Monday at GTC Taipei 2026, outlining the next evolution of AI compute.

Huang's presentation included updates on the Vera Rubin platform, a new lineup of Windows PCs developed with Microsoft for AI workflows, the launch of an enterprise agent toolkit, and next-generation AI infrastructure systems to accelerate data-center and agentic AI adoption.

A team of Goldman analysts, led by James Schneider, attended GTC Taipei 2026 and shared the top takeaways with clients.

Schneider had three key investment takeaways:

First, Nvidia (with Microsoft) is pursuing its traditional PC TAM more aggressively, which we believe could help drive some momentum for Windows on ARM (which has been extremely slow to date) given a concerted push with software partners.

Second, Nvidia continues to push its advantage in datacenter-level performance and cost leadership as a key differentiator relative to competitors - which we think should allow it to maintain competitive dominance at all but the largest hyperscalers.

Third, Nvidia is aggressively investing to drive the adoption of agentic AI across developers and ecosystem partners, and its Vera Rubin revenue ramp remains on track.

Here's more color on those takeaways:

Vera Rubin update: Nvidia announced that it is now ramping full production of its Vera Rubin platform, with multiple rack-scale systems (NVL72 GPUs, Vera CPUs, Groq 3 LPUs, BlueField storage, Spectrum-X networking) contributing to AI factory designs. The company highlighted that Vera GPUs are purpose-built for agentic AI use cases, with up to 1.8X the performance of X86 systems and 10X agent throughput vs. Blackwell. We expect a materially steeper revenue ramp for Rubin (beginning in 3Q) relative to Blackwell given meaningful manufacturing efficiencies and greater total capacity. In addition, the company highlighted its DSX AI Factory reference platform, which helps customers optimize their AI datacenters to bring operations up faster, while optimzing power consumption and system uptime.

New lineup of Windows PCs with Microsoft targeting AI workflows: Nvidia, in collaboration with Microsoft and Mediatek, launched a new Windows-based PC platform targeting AI workflows. The RTX Spark product combines a Blackwell RTX GPU with a 20-core Grace GPU (co-designed with Mediatek) using NVLink to deliver a high-performance PC experience optimized for AI applications - which we expect to be targeted at the premium segment of the market. OEM partners will launch laptop, desktop, and workstations systems beginning this fall, with launch partners including ASUS, Dell, HP, Lenovo, Microsoft, MSI, Acer and Gigabyte.

Launch of Enterprise Agent Toolkit. Nvidia announced a series of new software releases targeting agentic AI use cases in the enterprise, including NemoClaw, Nemotron 3 Ultra, OpenShell, and CUDA-X Agent Skills.

Physical AI announcements: Nvidia launched new versions of its open Cosmos (v3) frontier model targeting multi-modal reasoning, and Alpamayo (v2) which is targeted as a reference platform for self-driving cars. The company also announced its first open reference design for humanoid robots, based on its Isaac Gr00t and Jetson Thor hardware platform.

"The PC is being reinvented," Huang said. "For forty years, you launched apps. Click. Type. With RTX Spark and Microsoft Windows, you ask — and the PC does the work. RTX Spark brings everything NVIDIA has built — CUDA, RTX, our AI platform — into a single superchip. Local agents. Frontier models. Creative workflows. RTX games. All on a laptop. This is the new PC. The personal AI computer."

Nvidia shares rose 2.5% in premarket trading in New York after Huang's comments outlined that the company was entering the PC market with a new chip.

Arm ADRs soared 12% as traders viewed Nvidia's PC push as supportive of the Arm ecosystem. However, the announcement pressured incumbent processor stocks, with Intel sliding 6%, Qualcomm down 9.5%, and AMD falling 3.5%.

Schneider is "Buy" rated on NVDA with a 12-month price target of $285. This is based on a 30X P/E multiple applied to his team's normalized EPS estimate of $9.50. 

Meanwhile, overnight, Intel announced a new AI chip, code-named Crescent Island, expected to hit the consumer market by the end of the year, according to the Financial Times.

"We decided to start rebuilding our muscles in AI . . . [but] we are not particularly aiming for [the training market] based on past experience," said Kevork Kechichian, who leads Intel's data center group.

The AI chip race is accelerating, with today's biggest news being Nvidia's move to reinvent the PC market with a new AI chip.