Nvidia's Kyber NVL144 reportedly pushed back more than a year, Asian suppliers drop

2026-07-06 · Source: The Decoder · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Emerging Technologies & Innovation · Depth: Fundamental Awareness, quick

Summary

Nvidia's next-generation AI server system, Kyber NVL144, has been delayed to 2028, pushed back more than a year due to significant circuit board manufacturing problems, specifically with the PCB midplane. This delay, reported by SemiAnalysis, triggered sharp stock losses among key Asian suppliers, including Ibiden (down ten percent), Kingboard Laminates (down 18 percent), Elite Material (down ten percent), and Samsung Electro-Mechanics (down eleven percent). Furthermore, Nvidia has scrapped the planned NVL72x2 rack design and the more powerful four-die version of the Rubin Ultra chip, retaining only the two-die variant. A crucial interconnect technology, CPO-NVSwitch, is now deferred until the Feynman generation, creating a scaling gap for Rubin Ultra and potentially offering a timing advantage to competitors like AMD's MI500X and Google's TPUv8i Broadfly.

Key takeaway

For investors tracking AI infrastructure, Nvidia's Kyber NVL144 delay to 2028 and scrapped designs signal increased market volatility and potential shifts in competitive advantage. You should re-evaluate your portfolio's exposure to Nvidia's supply chain partners and consider the timing advantage this creates for alternative AI platforms from AMD or Google. This situation highlights the inherent risks in highly complex hardware roadmaps and the importance of manufacturing feasibility.

Key insights

Nvidia's advanced AI server and chip designs face significant delays and cancellations due to manufacturing and design challenges.

Principles

• Complex hardware development faces unforeseen manufacturing hurdles.
• Market reactions to tech delays can be swift and severe.
• Design complexity can lead to operational overhead issues.

In practice

• Monitor supply chain stability for advanced hardware.
• Evaluate competitor offerings during market shifts.
• Diversify hardware suppliers to mitigate risk.

Topics

• NVIDIA
• Kyber NVL144
• AI Servers
• Semiconductor Manufacturing
• Supply Chain
• Market Competition
• Rubin Ultra

Best for: CTO, VP of Engineering/Data, AI Architect, Investor, Director of AI/ML, Tech Journalist

Related on AIssential

• What to Expect from Nvidia This Week — Nvidia is expanding its AI hardware and ecosystem, while AI agents and models face both market adoption and regulatory challenges.
• Nvidia CEO Returns to South Korea as AI Memory Runs Short — AI industry growth is constrained by critical memory supply shortages, prompting direct executive intervention.
• AI chips are pushing everything else off TSMC's most advanced production lines — Simultaneous AI accelerator migration to TSMC's N3 process is creating a severe, multi-year capacity crunch.
• Amazon Wants to Sell Its Own AI Chips — Amazon's direct AI chip sales could disrupt Nvidia's dominance but face complex revenue and supply chain trade-offs.
• Why everyone from OpenAI to SpaceX is building their own chips (and turning up the heat on Nvidia) — Major tech firms are developing custom AI chips to reduce supplier dependence and optimize performance.
• Why GPU Server Prices Are Unpredictable — Limited component supply and high demand make GPU server prices unpredictable, impacting cloud provider costs and customer pricing.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Decoder.