BankofVol Grift
Publisher of Bank of Volatility, AI/ML Trading Bots
Copper connectivity stays in Nvidia's racks longer than anyone modeled, a win for Amphenol and Credo. Per @SemiAnalysis_, the Kyber NVL144 rack is delayed more than 12 months to 2028 because its PCB midplane cannot be made at scale, and the NVL72x2 backup was cancelled. Co-packaged optics will not arrive until the Feynman generation, so the shift away from copper in scale-up moves out with it. x.com
Beyond NVIDIA, TSMC and memory, the asymmetric plays in Vera Rubin NVL72's roughly $500B buildout are connector, power and cooling suppliers. Cableless racks raise content for Amphenol, TE Connectivity, Astera Labs, Credo and Aspeed. At 220kW a rack, power density is scarce, favoring Monolithic Power, Delta, Lite-On, Infineon, Vicor, Navitas and Allegro. Liquid cooling favors Vertiv, Schneider, nVent, Parker Hannifin, AVC, Auras and SPX; going chiller-less hurts Johnson Controls, Carrier, Trane. x.com x.com x.com x.com x.com
Credo is a main winner if the story @HyungsokAhn describes proves true, in which copper, pluggables, legacy power and testing benefit while co-packaged optics and 800VDC lose out. x.com
Credo Technology's copper interconnects should do well after SemiAnalysis reported that Nvidia's Kyber NVL144 rack has been delayed by more than 12 months to 2028 and its NVL72x2 back-to-back architecture was cancelled, leaving Rubin Ultra with a limited scale-up domain. x.com
Credo looks exposed as networking becomes AI's next bottleneck and Nvidia locks down optical fiber for scale-up, scale-out and scale-across links. MicroLED is readily available and a good contender for scale-up and potentially scale-out, with many Asian suppliers pivoting into it. x.com x.com
Copper scale-up is dead, making optics hyper bullish, especially Lumentum. Nvidia's Rubin NVLink Nyquist cut from 56 to 47 GHz signals failure: reflections matter, and bidirectional links make them worse. Feynman will have to use 16-wavelength DWDM ring-modulator CPO at roughly 50-64G NRZ. x.com x.com
Copper is not a bullish play, as it is arguably used more as a transitional tool over the next two years. That holds even though NVIDIA's CEO said they will need more copper, alongside optics and CPO. x.com
Semtech benefits from Nvidia's strategy of scaling up with copper for as long as possible. At Computex 2026, Jensen Huang said to use copper wherever you can and optics only where you must. Nvidia's Blackwell NVL72 racks use thousands of copper cables rather than fiber to link GPUs within a rack, and keeping data intact over that copper relies heavily on signal-integrity chips from Semtech's active copper cable alliance. x.com
Inference's move toward edge devices and dense, memory-hungry racks should drive demand for specialized chips, links, power delivery and cooling. McKinsey projects inference at 60% of AI computing demand by 2030. Ambarella makes on-device AI chips; Astera Labs links GPUs, memory and networks; Credo supplies short copper connections; Rambus earns from memory interfaces; Navitas reduces power waste; Modine cools dense racks; and Super Micro builds inference-specific server racks at volume. x.com … Show more
Copper gets a big win from the delay of Nvidia's Kyber NVL144 rack, which SemiAnalysis says has slipped more than 12 months to 2028 just three months after its GTC demo. Jensen Huang and Nvidia had hinted at the delay. x.com
Astera Labs and Credo Technology are the beneficiaries of copper scale-up interconnects outside Nvidia and AMD. Those two use proprietary scale-up protocols, while every other player's copper scale-up future lies with PCIe and UALink. x.com
Credo's interconnect opportunity grows as campus-scale AI clusters require more links to validate and integrate. The OCP 2026 meeting takeaway is that complex requirements put interconnect at the center, with both copper and optical links set to benefit. x.com
