SKYARK
ResearchGlobal macro perspective. Deep AI expertise. Independent research.
Research-Driven Conviction
Skyark Research is an investment research firm headquartered in Silicon Valley, managing approximately $1.5 billion in assets. Our work sits at the intersection of global macro analysis and deep technology sector expertise — with a particular focus on the artificial intelligence ecosystem.
Macro Orientation
We believe all assets are ultimately driven by macroeconomic forces — credit cycles, liquidity conditions, fiscal regimes, and the flow of global capital. Understanding the economic machine and where we sit within it is the foundation of every investment decision we make. This top-down lens shapes how we size positions, manage risk, and allocate across sectors.
AI Hardware Value Chain
Today, the clearest near-term opportunity in AI lies in the hardware layer. We cover the full stack — from GPU and accelerator design, advanced packaging and HBM memory, to optical networking, power delivery, and data center infrastructure. These are the physical bottlenecks and enablers of the AI buildout, and we believe the certainty of demand here is higher than anywhere else in the technology landscape.
AI Software & Applications
As the infrastructure matures, value will increasingly shift to the software and application layer. We are actively building vertical expertise in AI-native software, enterprise AI adoption, and the emerging platforms that will capture the next wave of monetization. This is a positioning effort — we are developing the research frameworks now to be ready when the inflection arrives.
Silicon Valley, California
The market is a weighing machine in the long run. Our work is to understand what it will weigh next — and to be positioned before the consensus arrives.Skyark Research
How We Think
Cyclical Thinking
Markets move in cycles — from broad economic expansions and contractions, to asset class rotations driven by shifting liquidity and risk appetite, down to the granular inventory and capex cycles within individual industries. We study these rhythms at every scale. Understanding where we are in a cycle — whether it is the credit cycle, the semiconductor cycle, or the AI infrastructure buildout cycle — is how we calibrate conviction, position sizing, and timing.
Deep Technical Understanding
We go beyond financial models to understand the technology itself — the architectures, the physics, the engineering tradeoffs. In the AI hardware stack, this means knowing how advanced packaging constrains GPU supply, where optical interconnects create bottlenecks, and which nodes in the supply chain capture disproportionate value. This technical depth allows us to identify where margin accrues, where disruption is real versus narrative, and where the next shift in value creation will emerge.
First-Principles Reasoning
We decompose every thesis into its fundamental assumptions — unit economics, total addressable demand, competitive moats, and capital intensity. When we evaluate a semiconductor company, we model wafer starts and yields, not just revenue estimates. When we assess a data center buildout, we calculate power requirements per rack and cooling costs per megawatt. This granularity is what separates durable conviction from surface-level pattern matching.
What We Study
Our research spans the verticals where we see the most compelling risk-reward — from the physical infrastructure enabling AI to the software platforms that will monetize it, and the frontier industries reshaping how we connect and explore.
Semiconductors & Compute
Coverage of the semiconductor value chain — GPU and accelerator architectures, advanced packaging (CoWoS, HBM), foundry capacity, EDA tooling, and the competitive dynamics among leading chipmakers. We track design wins, process node transitions, and supply-demand balances across the compute stack.
AI Infrastructure & Data Centers
Analysis of the physical infrastructure powering large-scale AI — hyperscaler capital expenditure, power and cooling systems, networking (optical interconnects, InfiniBand, custom Ethernet), and the emerging supply chain for custom silicon and next-generation accelerators.
AI Software & Platforms
Research into the software layer that sits on top of AI infrastructure — foundation model providers, enterprise AI platforms, vertical SaaS applications leveraging AI, and the developer tooling ecosystem. As compute scales, we are building the frameworks to evaluate where software monetization will concentrate.
Space & Frontier Technologies
Coverage of the commercial space economy — launch services, satellite communications, direct-to-device connectivity, and Earth observation. We study the convergence of declining launch costs, growing LEO constellations, and the expanding addressable market for space-based infrastructure and services.
Global Macro & Capital Flows
Top-down research on the macro environment shaping technology investment cycles. We study fiscal policy, interest rate regimes, trade dynamics, and cross-border capital flows as they relate to semiconductor supply chains, infrastructure spending, and the broader allocation of capital toward technology.
Built by Practitioners
Our team brings together professionals from top-tier hedge funds, global pension funds, venture capital firms, and leading technology companies. Members have held roles spanning technology equity research, semiconductor industry analysis, portfolio management, and early-stage investing — collectively contributing over five decades of experience across public and private markets.
Intellectual Rigor
Every investment thesis is stress-tested against data, alternative scenarios, and base-rate analysis before it informs a position.
Independent Judgment
We form our own views. Our value to investors comes from the willingness to disagree with consensus when the evidence supports it.
Continuous Iteration
Markets and technologies evolve. We update our models, revisit assumptions, and refine our frameworks as new information emerges.
Get in Touch
For institutional inquiries, research partnerships, or general information, please contact us directly.