Nvidia's Quiet AI Battles and Rising Competitors

Summary

Nvidia just announced its biggest earnings ever, with eighty-one point six billion dollars in revenue, an eighty-five percent year-over-year increase. However, a significant change in their reporting lumps gaming GPUs into a smaller edge computing segment, which now represents less than eight percent of their total revenue. This shift allows competitors like Qualcomm, Apple, ARM, AMD, and Intel to directly compare their stronger business units against Nvidia's less dominant ones, potentially impacting stock prices. In the data center market, Nvidia faces new competition from ARM and Qualcomm's data center CPUs, and from custom AI chips by companies like Google and Amazon for inference tasks. Qualcomm, for example, recently secured a deal to supply custom AI chips to ByteDance, the company behind TikTok, a market Nvidia cannot access due to export restrictions. Qualcomm's data center plans also include their Orion CPU and AI inference accelerators, contributing to their stock doubling in two months. Meanwhile, ARM is launching its AGI CPU, expected to generate over one billion dollars in its first year and potentially quadruple ARM's current annual revenue. Cerebras is also competing with its massive wafer-scale engine, which can process AI models significantly faster than Nvidia's chips on certain workloads. Despite these challenges, Nvidia's two-decade-old CUDA software platform remains a strong advantage. In the memory sector, Micron, the only US company producing high-bandwidth memory for AI data centers, reported record revenues of nearly twenty-four billion dollars last quarter, with nearly two hundred percent year-over-year growth and seventy-five percent gross margins. Their next quarter guidance projects forty percent revenue growth quarter-over-quarter. Lastly, the space market is poised for a major event with the expected IPO of SpaceX, potentially valued at two trillion dollars. Publicly traded Rocket Lab, a vertically integrated space company, launches small satellites with its three-dimensional printed engines and has secured two point two billion dollars in backlog, with two-thirds of its revenue from manufacturing spacecraft, not just launches. The AI revolution is extending into space, impacting both the chip and space industries.