Wall Street’s key takeaways from Nvidia CEO Jensen Huang’s speech and analyst Q&A at CES
Jensen Huang talks, Wall Street reacts.
When the leader of the biggest publicly traded company in the world delivers a 90-minute presentation, Wall Street is going to have some things to say about it.
Here’s what the sell side highlighted from Nvidia CEO Jensen Huang’s keynote address at the Consumer Electronics Show in Las Vegas on Monday, as well as the Q&A that the company held with analysts.
JPMorgan analyst Harlan Sur:
The confirmation that Vera Rubin chips are in full production and poised for a ramp in the second half of this year should be seen as “putting to rest recent concerns about potential delays.”
“NVDA has deftly positioned itself to benefit from multiple aspects of physical AI development — from data center compute (model training), to simulation (Omniverse) to edge devices (Jetson Thor) — which in aggregate could potentially drive the next leg of revenue growth for NVDA,” with Jensen Huang indicating that the AV revenue opportunity would be well above $10 billion by 2030.
The steady drumbeat of Nvidia championing the benefits of its GPUs relative to custom chips continues. The chip designer is “arguing that rivals (i.e. AI XPUs) will be hard pressed to keep pace with the performance of Vera Rubin and subsequent platforms with a ‘one chip at a time’ approach to design. We think to some degree there is validity to this notion, though growth in XPU volumes also clearly point to a TCO [total cost of ownership] benefit that hyperscalers are able to realize (in other words, AI ASICs continue to gain traction and capture share despite NVDA’s dominance in pure performance).”
It’s leaning in to inference-specific offerings: “NVDA also introduced a new context memory storage controller that aims to address the challenge of increasingly large context windows, opening up a new TAM [total addressable market] for the company where we think it will quickly gain traction with customers given the seamless integration of this storage system into its Vera Rubin platform offering.”
Morgan Stanley analyst Joseph Moore:
“Management described Rubin as being in ‘full production,’ highlighting meaningful improvements to manufacturability at the system level following their learnings with Blackwell. Rubin compute board assembly time has been reduced to ~5 minutes versus ~2 hours for Blackwell, and we saw on video the first rack being deployed. The timeline for launch remains second half of 2026, but revenue should be material around that time.”
“Management highlighted its unique position as the only chip company procuring these large quantities of DRAM and HBM directly, and with so much of the ecosystem supporting their growth they see themselves as having an advantage due to that scale.”
“No major surprises, but confidence on Rubin should be positively received given competitive noise exiting 2025 around broader TPU traction. While the obvious pushback is that Rubin specs and timelines haven’t changed, the stock is still 10% below highs immediately following Jensen’s $500 billion comments at GTC DC, numbers which have since moved higher post earnings and were reinforced in spirit today during the Q&A and fireside. With no hedging on supply or demand, we think enthusiasm can return as that plays out in numbers this year.”
Bank of America analyst Vivek Arya:
“AI scaling remains on track, with 5x token generation, 10x token cost reduction per year.”
Answering the where-does-the-money-come-from question: “AI to be funded by modernization of AI (repurposing $10 trillion of computing funding last decade), shifting of R&D methods.”
“China H200 demand is there, but still awaiting licenses”
Wedbush Securities analyst Dan Ives:
Overall, Huang’s address “fully set the tone for the AI Revolution heading into 2026 as the company laid the path for the next stage of the AI Revolution: physical AI.”
On Nvidia’s Alpamayo platform for autonomous vehicles: “We believe this new foundation model designed for AVs are an incremental positive for Tesla as the company looks to accelerate its autonomous vehicle technology and capitalize on the AV industry over the next decade.”
“We walked out of the event feeling even more bullish about Nvidia and the overall AI Revolution as the next stage of investments and technology are on the horizon that can facilitate a new age for the technology world with companies around the world are set to capitalize on $3 trillion to $4 trillion of AI Cap Ex hitting the market over the next 3 years.”
