Nvidia’s next AI server system, the Kyber NVL144, has been delayed to 2028 due to manufacturing difficulties. This shift caused immediate drops in share prices for Asian suppliers.
Analyst firm SemiAnalysis stated the setback stems from circuit board production. Specifically, the PCB midplane, a central board connecting individual components, has proven hard to make without defects. Nvidia CEO Jensen Huang demonstrated the rack just three months earlier at the company’s GTC conference.
Stocks slide
The report hit an already jittery investor base. After years of AI-fueled gains, even small setbacks trigger sharp sell-offs. Japanese PCB maker Ibiden, which counts Nvidia as its largest customer, dropped as much as ten percent. Kingboard Laminates fell 18 percent in Hong Kong. Elite Material lost ten percent in Taiwan. Samsung Electro-Mechanics slid eleven percent in South Korea.
The sell-off follows massive run-ups. Samsung Electro-Mechanics had gained more than 600 percent this year. Kingboard Laminates rose more than 470 percent.
Scrapped designs give competitors an opening
SemiAnalysis lists further setbacks. A planned alternative design called NVL72x2, which would have placed two Oberon racks back to back, has been scrapped entirely. Cloud providers and large data center operators pushed back against the unusual form factor and the high operational overhead. The more powerful version of the upcoming Rubin Ultra chip with four compute dies has also been canceled. Only the smaller two-die version remains, delivering roughly half the real-world compute performance.
A key interconnect technology called CPO-NVSwitch, which links many chips into a single large system, won’t arrive until the generation after next, called Feynman. That means Nvidia lacks a proven way to scale Rubin Ultra to very large systems for now. The gap could give competitors like AMD’s MI500X or Google’s TPUv8i Broadfly room to move in. Nvidia plans to make up for the shortfall by selling more Oberon-Rubin racks in the existing form factor.
A Kyber delay doesn’t necessarily mean overall AI spending will shrink, said Gary Tan of Allspring Global Investments. It shows that Nvidia’s most ambitious system is taking longer than expected. The current stock weakness is mostly driven by profit-taking. Shawn Oh of NH Investment & Securities pointed to growing uncertainty around Nvidia’s expansion plans, which gives alternative AI platforms more room to compete.
What it means
For teams building large-scale infrastructure, the timeline for the Rubin Ultra chip is now uncertain. The cancellation of the four-die version means fewer options for maximum performance in a single unit. Engineers must plan for a system that relies on more racks rather than fewer, larger ones. The delay in CPO-NVSwitch also means scaling will depend on older connections until the Feynman generation arrives.
