At the "Medical Spring Festival Gala," Nvidia gave a detailed talk on "How to do AI in healthcare."

At the recently concluded 44th Annual J.P. Morgan Healthcare Conference, NVIDIA once again demonstrated to the market how its computing power dominance is permeating the deepest parts of the real economy.

According to Hard AI, J.P. Morgan analyst Harlan Sur’s team released the latest research report on January 13, 2026, which thoroughly broke down NVIDIA Healthcare Vice President Kimberly Powell's speech. NVIDIA will leverage an "end-to-end" approach to transform the vast $4.9 trillion healthcare market into the next high-margin growth engine.

The report first points out the core logic of NVIDIA’s business model: the explosive profit margin driven by end-to-end vertical leverage. NVIDIA is building a closed loop from chips to tools to domain models, the so-called “dry lab -> wet lab” flywheel. For Wall Street, the most appealing narrative is that the same R&D platform can be reused indefinitely.

“Because the same core R&D platform can be reused horizontally (NVIDIA explicitly categorizes sovereign AI and enterprise AI as one, emphasizing the use of ‘the same tools’), over time, incremental victories in vertical domains will bring highly attractive operating leverage.”

Secondly, AI is moving from “pilots” to “employment.” 2025 is expected to be the breakout year for agentic AI with reasoning abilities, and now these digital employees are officially being put to work. J.P. Morgan observes that the speed of commercial AI deployment in the healthcare industry is three times that of the overall US economy, signifying a structural acceleration in enterprise AI adoption in this field. As inference costs have dropped more than 100-fold in the past four years, the ROI inflection point for large-scale adoption has arrived.

“NVIDIA is positioning itself at the platform layer, as spending shifts from pilot projects to paid deployment... Platforms like Abridge have already reclaimed more than 30% of clinicians’ time across more than 200 health systems worldwide.”

At the physical laboratory level, NVIDIA is attempting to eliminate humans as the “main data bottleneck” through cooperation with Thermo Fisher. By deploying the ‘three-computer platforms’ (COSMOS for simulation, Isaac for robot training, edge computing for deployment), NVIDIA is pushing for laboratory automation and intelligence.

“By directly integrating agent intelligence into instruments to automate experiment design and quality control... these autonomous labs can achieve a 100-fold increase in throughput and reduce production costs of complex drugs like cell therapies by 70%.”

Finally, the industrialization of drug discovery is accelerating. NVIDIA announced a milestone partnership with Eli Lilly, where both parties will jointly invest $1 billion over five years. This is more than a collaboration; it is a signal: to pharmaceutical giants, GPU clusters are no longer expendable IT spending, but critical production assets that determine survival.

“This signifies that GPU clusters are now regarded as essential capital infrastructure—akin to wet laboratories—that directly determine the success of the R&D pipeline.”

This article is from WeChat Official Account “Hard AI”. For more cutting-edge AI news, click here.

Risk Warning and DisclaimerThe market has risks; investment needs caution. This article does not constitute personal investment advice, nor does it take into account individual users' unique investment objectives, financial situation, or needs. Users should consider whether any opinions, views, or conclusions in this article suit their specific circumstances. Investments made accordingly are at one's own risk.