Cathie Wood is signaling a potential shift in the artificial intelligence chip landscape, as ARK Investment Management (ARK Invest) projects that Custom Silicon (ASICs) could significantly reshape competition between Nvidia Corp ($NVDA) and Amazon.com Inc. ($AMZN) by 2030.
While Nvidia remains the dominant force in AI accelerators today, ARK believes Amazon is emerging as a formidable contender, dubbing it the “Sleeping Giant.”
Meanwhile, NVIDIA Corporation (NVDA) closed at $177.19 on February 27, down 4.16%, and slipped another 1.47% in overnight trading to $174.59 as of 1:06 a.m. EST.
According to ARK Director of Research Frank Downing, this Custom Silicon trend could capture as much as one-third of the AI compute market by 2030, materially impacting Nvidia’s share.
ARK Director of Research Frank Downing wrote on X, “We predict that over a third of the compute market will be custom silicon by 2030. This announcement is another step in that direction. Everyone knows Google’s TPU, but Amazon is the sleeping giant that is waking up”. Cathie Wood reposted the post with the text: “Competition for Nvidia.”
NVIDIA’s Stronghold In AI Infrastructure
NVIDIA Corp has been the undisputed leader in AI accelerators, powering the majority of generative AI workloads. Its CUDA software ecosystem, developer loyalty, and rapid product cycles have built formidable competitive moats.
The company’s recent acquisition of Groq and continued innovation in AI inference and training hardware have further reinforced its leadership.
For now, Nvidia’s GPUs remain the gold standard for training and deploying large-scale AI models. Demand from hyperscalers, enterprises, and startups alike has driven record revenues and positioned $NVDA at the center of the AI infrastructure boom.
However, ARK’s research suggests that the next wave of AI infrastructure could be less dependent on general-purpose GPUs and increasingly reliant on tailored ASICs optimized for specific workloads. This is where Custom Silicon becomes the strategic bridge between Nvidia’s current dominance and emerging challengers like Amazon.
Amazon’s Custom Silicon Strategy
Amazon.com Inc. has been steadily expanding its semiconductor ambitions through AWS. The company’s Trainium chips, including Trainium 3 and the upcoming Trainium 4, are designed to optimize AI model training with improved cost efficiency and performance per watt compared to traditional GPU setups.
Amazon’s cloud dominance gives it a built-in distribution engine. Its partnership with OpenAI, alongside a record $200 billion capex in 2026, strengthens its ability to scale AI services while integrating Custom Silicon into its cloud offerings.
ARK reportedly argues that hyperscalers like Amazon can vertically integrate their compute stacks, reducing reliance on third-party chip vendors. As AI workloads grow more complex and widespread, the economics of internally designed ASICs could become increasingly attractive, particularly for high-volume inference tasks.
Downing described Amazon as the “Sleeping Giant” because it combined cloud scale, capital resources, and a chip development roadmap. If its Trainium chips continue to close the performance gap with Nvidia GPUs, the competitive landscape could shift meaningfully over the next several years.
The Broader Custom Silicon Movement
Amazon is not alone in developing proprietary AI chips. Alphabet Inc. (Google) has long deployed its Tensor Processing Units (TPU) to power AI workloads across Google Cloud and internal products. Meanwhile, Advanced Micro Devices, Inc. (AMD) continues to compete with Nvidia through its MI-series accelerators.
But ARK’s thesis emphasizes that large cloud providers have structural advantages when they internalize chip design. By controlling hardware, software, and cloud distribution simultaneously, companies like Amazon can optimize for cost, performance, and customer demand in ways traditional chipmakers cannot easily replicate.
The firm projects that by 2030, Custom Silicon could account for roughly one-third of AI compute demand, a substantial reallocation of market share within the semiconductor ecosystem.
What It Means For $NVDA And $AMZN Investors
For investors tracking $NVDA and $AMZN, ARK’s outlook introduces a longer-term competitive narrative rather than an immediate disruption. NVIDIA’s near-term fundamentals remain robust, supported by strong demand for AI accelerators and its entrenched developer ecosystem. Multiple architectures may coexist, with GPUs and ASICs serving different segments of the AI market.
Still, if ARK’s projections prove accurate, Amazon’s Custom Silicon ambitions could significantly reshape AI infrastructure economics by 2030.
