Why everyone from OpenAI to SpaceX is building their own chips (and turning up the heat on Nvidia)

A growing number of major technology companies, including OpenAI, Google, Apple, and SpaceX, are developing custom silicon chips to reduce their reliance on dominant supplier Nvidia and gain greater control over hardware performance [1]. Nvidia has long held a commanding position in the market for artificial intelligence processors [1]. The shift toward in-house chip design represents a strategic hedge rather than a complete severing of ties, with firms seeking hardware tuned to their specific computational workloads [1]. OpenAI disclosed its plans for a custom inference chip, codenamed Jalapeño, which it is building in partnership with Broadcom [1]. Broadcom is a major designer and supplier of semiconductors and infrastructure software, with its semiconductor products accounting for some 58% of its revenue as of 2025 [5]. The company has grown rapidly amid the AI boom, surpassing a $2 trillion market capitalization in April 2026 [5]. Apple was an early proponent of custom silicon, having moved its Mac computers away from Intel processors to its own designs, a transition that unlocked significant performance gains [1]. Apple is one of the Big Tech companies and has been the world's largest by market capitalization for most of the period since 2011 [2]. Google and SpaceX are also pursuing proprietary chip development, joining a widening field of firms investing in bespoke hardware to mitigate single-supplier risk [1]. The fabrication of these advanced custom chips will likely continue to depend on a small number of specialized manufacturers. Taiwan Semiconductor Manufacturing Company (TSMC) is the leading dedicated contract chipmaker, holding approximately 70% of the global semiconductor foundry market [6]. TSMC serves as the main supplier for Nvidia, Apple, Broadcom, and Qualcomm, and was the first foundry to commercialize extreme ultraviolet lithography technology in high volume [6]. Its central role in the production of advanced AI chips has raised concerns about supply chain bottlenecks [6]. Meta Platforms, another Big Tech firm, has also been described as part of the largest six U.S. technology companies by market capitalization, a group that includes Nvidia [4]. The push by multiple members of this cohort to design their own chips underscores a broader industry recalibration around supply chain resilience and hardware specialization [1].