New York, Jan 6, 2026, 10:26 EST — Regular session
- Nvidia up about 1.5% early Tuesday as CES chip roadmaps sharpen focus on 2026 AI hardware
- Nvidia targets second-half 2026 partner rollout for Rubin; Morgan Stanley flags a revenue ramp
- Next catalysts include Friday’s U.S. jobs report and Nvidia’s Feb. 25 earnings
Nvidia (NVDA.O) shares rose 1.5% to $190.99 in early New York trade on Tuesday, after the chipmaker laid out more of its next-generation Rubin computing platform for artificial intelligence (AI) at the CES tech show. The stock traded between $188.15 and $192.13, while rival Advanced Micro Devices (AMD.O) slid 1.5%.
Nvidia said Rubin is in full production and that products based on it will be available from partners in the second half of 2026. The company said the six-chip platform is designed to cut “inference” costs — the work of running trained AI models — by lowering cost per token (the chunks of text AI models process) and by training “mixture-of-experts” models, a specialist-heavy design, with four times fewer graphics processors (GPUs) than its Blackwell platform. OpenAI CEO Sam Altman said “Intelligence scales with compute,” in a statement backing the launch. NVIDIA Investor Relations
Joseph Moore at Morgan Stanley wrote that Rubin “will again raise the bar for performance” and said Nvidia’s CES message pointed to meaningful Rubin revenue in the second half of 2026. He also cited Nvidia’s claim that assembly time for a Rubin rack has been cut to about five minutes from roughly two hours for Blackwell. Investing
AMD CEO Lisa Su used CES to show MI455 data-center processors and unveil the MI440X, an enterprise chip aimed at on-premise deployments — running AI inside customers’ own server rooms rather than purpose-built clusters. OpenAI President Greg Brockman joined Su on stage and said chip advances were critical to OpenAI’s computing needs, while Su also previewed MI500 chips for 2027 that she said would deliver 1,000 times the performance of an older generation. Reuters
The chip moves came after a strong start to the year for U.S. stocks, with Monday’s rally lifting the Dow to a record close as investors rotated into financials. The next test for risk appetite is Friday’s U.S. nonfarm payrolls report, which traders see as a key read on Federal Reserve policy; markets are pricing about 60 basis points of rate cuts in 2026, LSEG data showed. Reuters
Rubin will land into a tighter race for the inference market, where AMD and Alphabet’s Google are pushing their own chips and systems to run chatbots at scale, and Nvidia’s CEO said Rubin’s performance step depends on a proprietary data format Nvidia wants the industry to adopt. Nvidia also described a “context memory storage” layer — a storage approach for long chat histories — and said its co-packaged optics networking technology, which uses optical links inside switches, will compete with offerings from Broadcom and Cisco. Nvidia executives said demand for its current H200 chips in China remains strong, but the company is waiting for export licenses to ship them. Reuters
Nvidia is scheduled to report fourth-quarter fiscal 2026 results on Feb. 25, and investors will watch for guidance on supply, pricing and the pace of Rubin adoption. NVIDIA Investor Relations
