NVDA Smashes Q1 Guide: What It Means for Micron, Memory, and the AI Trade

About this episode

Welcome back to StocktwitsTV on the road. Host Michele Steele is joined by Michael Parekh, ex Goldman and author of the I Return to Zero Substack, to break down Nvidia earnings as the numbers hit.

Nvidia delivered a massive quarter with revenue coming in at $68 billion and record data center revenue of $62 billion. The market immediately shifted to the April quarter guide, where Nvidia forecast Q1 revenue of $76.4 to $79.5 billion, beating expectations by several billion dollars. Michael’s takeaway is simple: the AI infrastructure ramp is not slowing, it is compounding, cutting through a wall of worry around bubbles, demand, and the broader AI narrative.

They dig into what is actually constrained in this cycle. Michael says the entire AI data center stack is in short supply for the next year or two, from GPUs to memory chips, and he flags power as another critical input the tech industry does not fully control. Michele asks about gross margins, which came in at 75 percent, and Michael explains why Nvidia, like Apple, sits at the front of the line at Taiwan Semiconductor, giving it more flexibility than competitors, even as demand remains broad enough that multiple players can benefit.

Finally, they discuss how this beat could impact the broader sector. Michael expects sentiment to stay volatile but rejects the idea of a software apocalypse, aligning with Jensen that the selloff logic is wrong. He highlights Jensen’s message that computing demand is growing exponentially and that the agentic AI inflection has arrived, pointing to real traction in tools like Claude Code and newer open agent frameworks that expand what chips can enable across industries.

Disclaimer: All opinions expressed on this show are solely the opinions of the hosts’ and guests’ and do not reflect the opinions of Stocktwits, Inc. or its affiliates. The hosts are not SEC or FINRA registered advisors or professionals. The content of this show is for educational and entertainment purposes only. Please consult with your financial advisor before making any investment decision. Read the full terms & conditions here: https://stocktwits.com/about/legal/terms/

Chapters and Timestamps

00:00 Intro
00:58 Nvidia quarter: $68B revenue, $62B data center record
01:32 Q1 guide: $76.4 to $79.5B and what it signals
02:18 Demand “off the charts and sold out” plus memory stack impacts
02:41 Power as the other key bottleneck
03:04 Gross margin 75 percent: peak or durable
03:25 Supply chain advantage at Taiwan Semiconductor
03:44 Competitors and why it is not zero sum
04:08 Hyperscaler CapEx and diminishing returns debate
05:03 Why GPUs still take the bulk of spend, memory rising but smaller on data centers
05:51 NVDA after hours reaction
06:12 Demand runway: limited for 1 to 3 years, secular turn
07:13 Chatbots to agents: why compute needs go 5x to 10x
08:08 Valuation ceiling question and the post Blackwell roadmap
09:15 Pickaxes and shovels: why investors extrapolate to 2030
09:58 Jensen and the software selloff: “illogical”
10:26 Volatility and the wall of worry
11:10 Why the software apocalypse thesis is wrong
11:55 Bottom up understanding: what these tools enable
12:23 Jensen quote: computing demand exponential, agentic inflection arrived
12:42 Why agentic is now practical
13:08 Claude Code traction and exponential adoption
13:56 Why Jensen’s words carry more weight
14:30 Open source agents and OpenAI acquisition mention
14:51 Wrap and Michael Parekh plug