Last year I wrote two in-depth articles about Nvidia to conclude. When I wrote “What’s Next for Nvidia Stock in 2026,” I thought there wouldn’t be any news about Nvidia for a while.
Then, NVIDIA completed its investment in Intel, and I wrote “NVIDIA Delivers on Key 2025 Commitments” to analyze this cooperation in detail.
I thought January would be quiet then, but I was wrong.
Nvidia decided to go all out with its announcement during CES. It looks like the company has nothing more to reveal at the GTC conference in March.
The company even unveiled its next-generation GPU, Vera Rubin, something it would normally reserve for the GTC.
Why so anxious? Is the bubble about to burst, or is AMD’s Helios rack system, which will also be launched this year, putting pressure on Nvidia?
I think it was AMD’s Helios and Google’s TPU that transformed Nvidia, and that’s reflected.
Nvidia CEO Jensen Huang believes “the ChatGPT moment for physical AI has arrived.”Shutterstock – Glenn Photos
Bank of America analyst Vivek Arya and his team participated in the Nvidia (NVDA) CES 2026 keynote speech and financial analyst Q&A session on January 5. Following the event, they updated their view on NVDA stock in a research note shared with TheStreet.
Nvidia CEO Jensen Huang noted that demand for AI computing continues to be “very high” and announced the launch of the new Vera Rubin AI platform, the team said.
-
AI scaling remains on track, with annual token generation increasing fivefold and costs decreasing tenfold.
-
Announced the launch of six new AI chips for the Vera Rubin platform, expected to be launched in the second half of 2026.
-
The company has launched a new pod-level contextual memory storage platform.
-
Today, NVIDIA continues to operate every major LL.M.
-
Funding for AI will come from the modernization of AI and changes in research and development methods.
-
Groq/SRAM transactions may be beneficial for very low-latency workloads.
-
Artificial intelligence is moving beyond the LL.M. and into the field of physical artificial intelligence.
-
The demand for H200 in China is there but still waiting for permission.
According to CNBC, informed and attentive readers will notice that the analysis team must have overlooked the fact that Google Gemini 3 is trained and run on Google’s own TPU.
The team said that despite Nvidia’s earnings per share compound annual growth rate of more than 35% and free cash flow of more than 40%, its continued dominance in artificial intelligence computing, network systems and ecosystems is trading at a price-to-earnings ratio of only about 19 times, or in line with the broader S&P 500 index.
Arya reiterated a Buy rating and $275 price target based on his 2027 price-to-earnings estimate of 28 times excluding cash, which is within Nvidia’s historical forward P/E range of 25 to 56.