New York, Feb 3, 2026, 06:10 EST — Premarket.
- Nvidia slipped in premarket trading following reports that OpenAI is dissatisfied with some of its latest chips and is looking into other options.
- AMD climbed before earnings, as investors sought new clues on data-center demand ahead of the after-hours report.
- Palantir and Teradyne rose following positive news linked to AI investment.
Nvidia shares dipped 2.9% to $185.61 in premarket trading Tuesday following reports that OpenAI is exploring alternatives to several of the chipmaker’s newest AI processors. (Reuters)
The news comes at a tricky time for the AI sector. Investors are weighing if the boom is moving from training massive models to “inference”—the stage where models produce answers for users—and what that means for chip makers. (Reuters)
The next move hinges on earnings and guidance rolling out over the next day. AMD reports after the bell. Server specialist Super Micro Computer will update investors Tuesday. Traders are also parsing fresh numbers and outlooks from Palantir and Teradyne. (Yahoo Finance)
Reuters reported late Monday that OpenAI has been unhappy with the performance of certain Nvidia chips on inference-heavy tasks and has been searching for alternatives since last year, according to eight sources familiar with the issue. The company has looked into options like AMD and startups Cerebras and Groq. (Reuters)
Nvidia said customers “continue to choose NVIDIA for inference” thanks to its performance and “total cost of ownership at scale.” An OpenAI spokesperson confirmed the company relies on Nvidia to power the “vast majority” of its inference fleet. OpenAI CEO Sam Altman called Nvidia’s chips “the best AI chips in the world” and said OpenAI aims to remain a “gigantic customer.” (Reuters)
A Reuters report highlighted a battle over memory and speed, noting that OpenAI is exploring chips with substantial SRAM — memory built directly onto the chip — since inference often hits a bottleneck on memory access, not just raw computing power. (Reuters)
On Monday, OpenAI took another step into coding by rolling out a desktop app for Codex, aiming to carve out a bigger slice of the code-generation market. Altman highlighted the appeal of these tools, saying, “They keep trying, they don’t run out of motivation.” (Reuters)
AMD shares climbed 4.0% to $246.27 ahead of Tuesday’s open, with investors gearing up for the company’s quarterly earnings report. Wall Street is focused on any changes in forecasts for data-center demand and AI accelerator sales, given the recent surge that’s put extra weight on this quarter’s guidance. (Yahoo Finance)
Super Micro Computer climbed 2.1% to $29.71 in premarket trading before its fiscal second-quarter earnings call set for 5:00 p.m. ET, the company said in a statement. (Supermicro)
Palantir’s stock climbed 0.8% to $147.76 following a quarterly sales report of $1.41 billion. U.S. government revenue surged 66% to $570 million. CEO Alex Karp stood by the company’s surveillance technology, emphasizing that oversight features are built into its platform. (Reuters)
Valuation is still a hot topic for the stock. eToro analyst Zavier Wong noted that “valuation question marks won’t disappear,” pointing out that Palantir “remains priced for perfection.” (Reuters)
Teradyne jumped 3.6% to $249.53 in premarket action after predicting quarterly revenue and earnings that topped Wall Street’s forecasts, fueled by AI-fueled demand for semiconductor test equipment. CEO Greg Smith highlighted “strong momentum in compute driven by AI.” (Reuters)
A major “but” looms over the story. Even if OpenAI brings on new suppliers, Reuters noted Nvidia still dominates the chips used to train large AI models. OpenAI itself emphasized that Nvidia powers the bulk of its inference fleet — highlighting how switching costs and software tools can stall shifts in market share. (Reuters)
Attention is turning to Tuesday’s key events: AMD will release earnings after the market closes, followed by Super Micro’s 5 p.m. ET earnings call. Traders will also be watching for any updates on OpenAI’s chip plans and the progress of its Nvidia-related investment talks. (Supermicro)
