Key Takeaways
- • Grayscale Investments launched Grayscale AI Compute ETF (ticker: GCPU) on September 22, 2026, tracking the Indxx High Performance Computing Index to provide equity exposure to global companies building AI infrastructure.
- • GCPU targets a $1 trillion+ annual AI-related capital expenditure market, with the majority allocated to physical infrastructure, while data centers operate at record-low vacancy rates with only approximately six months of available capacity in reserve.
- • The ETF captures two distinct growth paths: newly constructed data center infrastructure and the repurposing of existing power, land, and grid-connected facilities (including Bitcoin mining operations) for AI compute workloads.
- • GCPU trades at approximately $38.54–$39.76 per share with a current dividend yield of 0.14%, offering institutional investors diversified exposure to the infrastructure gap between AI's compute demand and available physical supply.
- • Grayscale leverages its 13-year track record in digital asset investing to position compute infrastructure as a scarce, transformational technology category alongside prior asset classes.
Summary
Grayscale Investments, the world's largest digital asset-focused investment platform, introduced the Grayscale AI Compute ETF (GCPU) on September 22, 2026, marking an expansion of its investment thesis into physical infrastructure underpinning artificial intelligence growth. The ETF is designed to track the Indxx High Performance Computing Index and offers equity exposure to a global basket of companies engaged in building and operating the data centers, power systems, and compute capacity required to meet surging AI demand.
The investment case centers on a structural supply-demand imbalance in AI infrastructure. Data centers across North America are operating at record-low vacancy rates, with approximately six months of available capacity in reserve, while new capacity typically requires two to five years to bring online. Industry projections indicate annual AI-related capital expenditure will exceed $1 trillion, with the majority directed toward physical infrastructure. This growing gap between AI's compute demand and available physical supply represents what Grayscale characterizes as a durable and investable dynamic within the broader AI theme.
GCPU pursues two distinct pathways to AI compute capacity growth: newly built data center infrastructure and the repurposing of existing power, land, and grid-connected facilities for AI workloads. The latter category includes Bitcoin miners and other legacy industrial operators transitioning assets to support AI compute. According to Steve Vanourny, Head of Index at Grayscale, "GCPU is a direct extension of that thesis, giving investors access to the companies closing the gap between AI's demand for compute and the physical infrastructure needed to meet it."
The ETF reflects Grayscale's 13-year track record in digital asset investing innovation and positions compute infrastructure alongside prior scarce, transformational technologies. At launch, GCPU traded at approximately $38.54–$39.76 per share with a dividend yield of 0.14%, providing institutional investors with systematic exposure to companies addressing the critical infrastructure bottleneck in AI's continued expansion.
The investment case centers on a structural supply-demand imbalance in AI infrastructure. Data centers across North America are operating at record-low vacancy rates, with approximately six months of available capacity in reserve, while new capacity typically requires two to five years to bring online. Industry projections indicate annual AI-related capital expenditure will exceed $1 trillion, with the majority directed toward physical infrastructure. This growing gap between AI's compute demand and available physical supply represents what Grayscale characterizes as a durable and investable dynamic within the broader AI theme.
GCPU pursues two distinct pathways to AI compute capacity growth: newly built data center infrastructure and the repurposing of existing power, land, and grid-connected facilities for AI workloads. The latter category includes Bitcoin miners and other legacy industrial operators transitioning assets to support AI compute. According to Steve Vanourny, Head of Index at Grayscale, "GCPU is a direct extension of that thesis, giving investors access to the companies closing the gap between AI's demand for compute and the physical infrastructure needed to meet it."
The ETF reflects Grayscale's 13-year track record in digital asset investing innovation and positions compute infrastructure alongside prior scarce, transformational technologies. At launch, GCPU traded at approximately $38.54–$39.76 per share with a dividend yield of 0.14%, providing institutional investors with systematic exposure to companies addressing the critical infrastructure bottleneck in AI's continued expansion.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.