Nvidia's AI Capex Sustainability: TSMC's Packaging Limits and the Kyber NVL144 78-Layer PCB Failure

Updated

Nvidia's AI Capex Sustainability: TSMC's Packaging Limits and the Kyber NVL144 78-Layer PCB Failure

The structural demand driving Nvidia’s artificial intelligence dominance remains robustly supported by physical supply chain signals, but its breakneck annual release cadence is hitting severe physical and manufacturing bottlenecks. In July 2026, the narrative around Nvidia's upcoming next-generation architectures—specifically the Vera Rubin and Rubin Ultra platforms planned for 2027—has split between detailed supply chain reports of major delays and firm management denials.

The Kyber NVL144 78-Layer PCB Manufacturing Bottleneck

On July 6, 2026, semiconductor research firm SemiAnalysis reported that Nvidia's next-generation Kyber NVL144 rack-scale AI system—designed to house 144 Rubin Ultra GPUs in a single unified compute domain—has been delayed by more than 12 months to 2028.

The source of the delay is a manufacturing failure at the absolute physical frontier of printed circuit board (PCB) fabrication. Unlike the current Vera Rubin NVL72 systems, which rely on approximately 5,000 copper cables totaling nearly two miles in length to link 72 GPUs, the Kyber architecture was designed to eliminate cables using an orthogonal backplane PCB. This board sits between vertically mounted compute trays and rear switch trays, routing all NVLink signals through rigid copper traces at SerDes speeds exceeding 448 gigabits per second per lane.

However, the physical specifications of this orthogonal backplane have proven too complex for current manufacturing capabilities:

  • 78-Layer Stack: The board is roughly the size of a kitchen table and requires laminating three separate 26-layer circuit boards together under extreme heat and pressure.
  • Specialty Materials: The stack uses a hybrid of M9-grade copper-clad laminate, quartz fabric, and PTFE.
  • Extreme Precision: Copper traces must be held to widths and spacings of 25 micrometers or less, with impedance controlled within a ±5% tolerance across the entire board.

According to CITIC Securities, this combination of extra-large form factors and extremely high layer counts presents challenges "significantly greater than conventional products in areas such as yield control, impedance consistency, and thermal management." No advanced PCB fabricator has been able to produce this board at acceptable production yields.

Rejected Backup Plans and Rubin Ultra Spec Downgrades

To bridge the gap, Nvidia proposed a contingency design called the NVL72x2, which would place two existing Oberon NVL72 racks back-to-back and link them via copper NVLink cables. However, cloud service providers (CSPs) and hyperscalers heavily rejected the design due to its odd configuration and heavy operational and cabling burdens, forcing Nvidia to cancel the backup plan.

This delay follows another major setback reported on June 30, 2026: the quiet cancellation of the original quad-chiplet version of the Rubin Ultra GPU. Due to advanced packaging limits at TSMC, Nvidia had to abandon the quad-chiplet design (which would have carried 16 HBM4E memory stacks for 1 terabyte of memory per package). Instead, Nvidia will ship a dual-chiplet version carrying roughly 8 HBM4E stacks, effectively halving the planned memory capacity.

Nvidia's Response and Near-Term Financial Strengths

Nvidia has swiftly and urgently denied the delay claims. On July 15, 2026, CEO Jensen Huang dismissed reports of delays in the Vera Rubin rollout, stating that the next-generation systems are "in production and on track" and affirming "giant" production volumes. The company's official response remains: "Our roadmap is intact."

Analysts note that these roadmap issues do not affect Nvidia's immediate financial trajectory:

  • NVL72 Production: The current Vera Rubin NVL72 systems—built on the existing Oberon rack architecture, which is entirely unaffected by the Kyber midplane issue—are in full production and scheduled for delivery this fall (2026) to eight confirmed cloud partners (AWS, Microsoft Azure, Google Cloud, Oracle Cloud, CoreWeave, Lambda, Nebius, and Nscale).
  • Strong Financial Performance: In the quarter ending April 30, 2026, Nvidia reported a record $81.61 billion in revenue (+85.2% YoY) and $58.32 billion in net income, supported by a 74.1% gross margin and 63.0% profit margin (/markets/NVDA/2026/07/21).
  • Valuation Support: Even with recent volatility, Nvidia's forward P/E multiple sits at a reasonable 21.7x, close to the S&P 500 average and well below its five-year average of 72x, while SemiAnalysis projects Nvidia's second-half fiscal 2027 data-center compute revenue will exceed Wall Street consensus estimates by 20%.

The high-stakes test of who is correct will arrive on August 26, 2026, when Nvidia reports Q2 FY2027 earnings and management is forced to address the Kyber timeline and PCB yields on the record.

Part of

This finding is an example of a pattern recurring across your work:

Backlinks

Revision history

  • Update Nvidia's AI capex sustainability finding with the detailed Kyber NVL144 78-layer PCB manufacturing delay, the Rubin Ultra memory spec downgrade, and Jensen Huang's July 15, 2026 denial.
    · by the agent
  • Update the Nvidia capex sustainability note with the technical details of the July 6, 2026 Kyber NVL144 delay, the 78-layer PCB manufacturing failure, the cancellation of the NVL72x2 backup plan, and the dual-chiplet Rubin Ultra modification.
    · by the agent
  • Update Nvidia's AI capex sustainability note with the recent Kyber NVL144 delay to 2028, Rubin Ultra redesign, and New York's historic data center moratorium.
    · by the agent
  • Update Nvidia's AI capex sustainability finding with TSMC's blockbuster Q2 2026 earnings (capex upgraded to $64B) and the SemiAnalysis disclosure of the Kyber NVL144 rack delay to 2028 due to PCB midplane manufacturing bottlenecks.
    · by the agent
  • Update Nvidia AI capex sustainability note with TSMC's blockbuster Q2 2026 earnings call results, its $60B-$64B capex hike, the $100B Arizona expansion, and the agentic AI CPU resurgence theme.
    · by the agent
  • Update the Nvidia AI Capex sustainability note with TSMC's blockbuster Q2 2026 earnings results and upgraded CapEx/revenue guidance released on July 16, 2026.
    · by the agent
  • Update the Nvidia AI capex sustainability note with TSMC's blowout June/Q2 2026 revenue, Goldman's $5.3T hyperscaler capex projections, and the SemiAnalysis confirmation of the Rubin Ultra four-die package cancellation due to CoWoS-L warpage limits.
    · by the agent
  • Update the Nvidia AI capex sustainability note with TSMC's blowout June/Q2 2026 revenue, Goldman's $5.3T hyperscaler capex projections, and the SemiAnalysis confirmation of the Rubin Ultra four-die package cancellation due to CoWoS-L warpage limits.
    · by the agent
  • Update the Nvidia capex cycle sustainability note with Q1 FY2027 footnote disclosures, the July 2026 Kyber rack delay debate, and record TSMC Q2 sales ahead of the July 16, 2026 conference.
    · by the agent
  • Update Nvidia's AI capex sustainability note with breaking Q2/June sales data from TSMC and Ornn B200 spot rental price spikes.
    · by the agent
  • Update Nvidia's capex sustainability story to include the Kyber delay denial, Goldman's compelling valuation call, Morgan Stanley's broader AI demand note, and Supermicro's Rubin NVL4 blueprint.
    · by the agent
  • Update Nvidia's AI capex sustainability note with detailed engineering analysis of the 78-layer PCB midplane, exotic materials (M9 CCL, Quartz glass), the cancellation of the NVL72x2 back-to-back rack design, TSMC's CoWoS capacity status, and upcoming earnings catalysts.
    · by the agent
  • Update Nvidia's AI capex sustainability note with the July 2026 Kyber NVL144 rack delay controversy, Nvidia's official denial, and latest NVDA market metrics.
    · by the agent
  • Update the Nvidia AI capex sustainability note with fresh July 2026 data, including SemiAnalysis' $11.1T capex projections, TSMC's anticipated $56B/$73B capex guidance, and Samsung's blowout Q2 2026 preliminary operating profit of 89.4T won.
    · by the agent
  • Update Nvidia's AI capex sustainability note with July 2026 cloud capex projections ($1T in 2026, $1.5T in 2027), supply constraints, and semiconductor ecosystem earnings (Samsung, Broadcom, Micron).
    · by the agent
  • Update Nvidia's AI capex sustainability story with early July 2026 data, including the SemiAnalysis Kyber NVL144 delay report, Nvidia's official denial, and JPMorgan's raised $5.5T global AI capex projections.
    · by the agent
  • Update Nvidia AI capex sustainability with record Q2 2026 Foxconn (Hon Hai) sales beating estimates and JPMorgan raising long-term global AI capex projections to $5.5 trillion.
    · by the agent
  • Update the capex sustainability finding to incorporate the Rubin platform delay risks and the subsequent extension of Blackwell's shipment dominance (>70% in 2026).
    · by the agent
  • Update the AI capex sustainability debate with Nvidia's latest Q1 FY2027 market view, Ornn Index Blackwell rental rates surging 114% to $4.95/hr, and TrendForce's June 2026 report on CoWoS supply-demand gap narrowing.
    · by the agent
  • Update the Nvidia AI capex sustainability note with Microsoft's $190B capex, CoreWeave's $31B-$35B capex, Nebius' $20B-$25B capex, and Ornn Index Blackwell rental rates rising to $4.08-$4.95/hr.
    · by the agent