Nvidia Corp (NASDAQ:NVDA) CEO Jensen Huang said he is "100% confident" that OpenAI, Anthropic and other frontier AI labs will remain Nvidia customers and partners "for a very long time," despite investing billions in custom silicon. Huang said many in-house processors target narrow inference workloads for a single cloud or service, while Nvidia provides an end-to-end AI platform spanning the full AI lifecycle and expects AI services to continue running on its infrastructure worldwide. His comments followed OpenAI's benchmark results for Jalapeño, a custom AI inference chip developed with Broadcom Inc (AVGO), which OpenAI said delivered more AI work per unit of power and faster responses than Nvidia's GB300 in testing. The comparison did not include Nvidia's newer Vera Rubin platform, and Jalapeño is not designed for AI training. Huang also said Nvidia's investments in frontier AI companies represented a "once in a generation opportunity" and that his only regret was not investing more and sooner. Nvidia reported Q2 FY2026 revenue of $96.22 billion, up 106% from a year earlier and above the $92.18 billion Street consensus, with gross margin of 75%. The company forecast third-quarter revenue of $105.84 billion to $110.16 billion, up 85.7% to 93.2% year over year.