Nvidia Hits Record Highs On CES AI Breakthroughs: Goldman Stays Bullish

Nvidia Corporation NVDA saw its stock price climb to fresh all-time highs in Tuesday's premarket trading following its show-stopping presentation at CES 2025 in Las Vegas, where the company unveiled a bold vision for the future of artificial intelligence (AI) and robotics.

Founder and CEO Jensen Huang delivered a keynote Monday that emphasized Physical AI as the next frontier for the robotics industry, likening the moment to what ChatGPT represented for natural language processing.

In a note shared Tuesday, Goldman Sachs analyst Toshiya Hari reiterated his Buy rating on Nvidia stock, which is also on the firm's Conviction List.

"We believe today's string of announcements, at a minimum, highlights the company's ability to innovate at industry-leading speed across hardware and software as well as its robust partner and customer ecosystem," Hari said.

A New Standard In Gaming And Computing Power

Central to Nvidia's CES announcements was the launch of its RTX Blackwell family of GPUs, which includes the GeForce RTX 5090 and RTX 5070.

These next-generation GPUs are poised to redefine gaming and creative computing, offering unprecedented performance gains.

The RTX 5090, which features 92 billion transistors and 3,352 AI TOPS of computing power, is designed to deliver 2x the performance of the RTX 4090, thanks to AI-driven rendering technology. This advancement allows the GPU to generate three frames for every one rendered, dramatically boosting frame rates.

For budget-conscious gamers, Nvidia introduced the RTX 5070 at $549, which promises performance equivalent to the current RTX 4090, a GPU priced at $1,599.

According to Huang, this family of GPUs demonstrates Nvidia's focus on driving down the cost of compute as models, including OpenAI's o1 and o3 and Google's Gemini Pro, continue to increase in complexity.

Physical AI: The “Largest Technology Industry the World's Ever Seen”

Huang's keynote went beyond GPUs, focusing heavily on Physical AI—a domain encompassing robotics, autonomous systems, and real-world AI applications.

According to Hari, Nvidia is positioning itself to lead what Huang described as "the largest technology industry the world's ever seen."

Nvidia’s strategy is grounded in three key scaling laws:

1. Pre-training scaling, which improves AI models by applying more computational power to larger datasets.
2. Post-training scaling, which enhances models through reinforcement learning.
3. Test-time scaling, where models gain reasoning and problem-solving capabilities.

Nvidia Unveils ‘Cosmos’

Nvidia introduced Cosmos, an open platform designed to democratize physical AI development for applications like autonomous vehicles and humanoid robots.

Hari said, "Cosmos (which NVDA has made available under an open model license) democratizes the development process through synthetic data for training and evaluation."

The platform also integrates with Nvidia's Omniverse, enabling developers to simulate AI scenarios—such as snowy roads or crowded warehouses—without relying on real-world data collection.

Llama Nemotron And The Future Of Robotics

In collaboration with Meta Platforms Inc. META, Nvidia unveiled the Llama Nemotron Language Foundation Models, designed for enterprise AI applications such as fraud detection and customer support.

Hari highlighted, "These models are optimized for Agentic AI for enterprises and provide options for different scale deployments."

Hari indicated that the development of humanoid robots hinges on processing imitation information, which Nvidia aims to accelerate using its Omniverse platform.

Market Reactions And Goldman Sachs’ Nvidia Outlook

Shares of Nvidia were up 2.3% to $152 during Tuesday’s premarket trading , poised to open the session at record highs.

Goldman Sachs holds a 12-month price target of $165 on Nvidia, based on a 50x multiple of its normalized EPS estimate of $3.30.

Read now:

• AMD Unveils ‘Next Generation’ AI Processors, Gaming Gear Ahead Of CES 2025

Photo: Shutterstock