In the semiconductor industry, the decisions made in executive suites do not just shape quarterly earnings — they determine whether a company leads the next wave of computing or falls behind it. Against that backdrop, the CEO compensation packages at Nvidia (NVDA), Advanced Micro Devices (AMD), and Intel (INTC) between 2021 and 2024 offer compensation professionals a rare, real-world laboratory for studying pay-for-performance alignment in action.
The results are striking. Where compensation structures tracked closely with long-term corporate strategy, companies thrived. Where they diverged — or where incentive design failed to account for the complexity of a turnaround situation — the consequences proved costly. Here is what the data shows, and what it means for how we think about executive reward design.
The Structure Tells the Story
All three companies publicly espouse a pay-for-performance (PFP) philosophy — and all three rely heavily on equity as the dominant compensation vehicle. But the architecture of those equity awards diverges meaningfully, and those divergences carry a real strategic signal.
At Nvidia, the design is aggressive by intent, and it has worked. When performance delivers, the compensation system rewards. When it does not — Huang received no cash bonus in 2023 when targets were not met — it doesn't. At AMD, the diversified incentive design mirrors AMD's strategic posture: innovate steadily, build market share, and reward patience. At Intel, the structure failed not because the metrics were wrong, but because the design did not create sufficient urgency proportionate to the challenge being faced.
"A compensation structure is not just a retention tool — it is a directional signal about what the board believes will create value, and on what timeline."
Pay-for-Performance: A Theory Stress-Tested
Agency theory — the long-standing academic framework that underlies most executive compensation design — holds that aligning executive financial interests with shareholder outcomes will produce the right behaviors. Nvidia and AMD largely validate this premise. Huang's wealth is inextricably bound to Nvidia's market performance; Su's stock options become worthless if AMD's stock stagnates. Both CEOs are structurally incentivized to act like owners.
Intel's case complicates the picture. Gelsinger's compensation was designed with the right metrics in principle — revenue, cash flow, and TSR — but the execution fell short. A 1,171:1 CEO-to-median-worker-pay ratio in 2021, driven by a $140 million new-hire equity grant, triggered shareholder backlash and a say-on-pay vote against the company's executive compensation program. It was a signal that compensation design, however well-intentioned, must also contend with legitimacy in the eyes of the broader stakeholder audience.
This is where stewardship and stakeholder theories add a dimension that pure agency thinking misses. The multi-year vesting schedules embedded in all three packages — four years for Huang's single-year PSUs, three years for Su's PRSUs — represent an attempt to institutionalize a longer-term orientation. Whether they succeed depends heavily on the culture and expectations that surround them.
The ESG and Stakeholder Gap
One of the most notable findings across all three compensation structures is the near-total absence of non-financial performance metrics. TSR, revenue growth, operating income, and EPS dominate every equity award design. Environmental, social, and governance (ESG) metrics — employee satisfaction, carbon reduction, supply chain equity, community investment — are conspicuously absent, even as all three companies operate at the intersection of national security, global supply chains, and transformative AI development.
Research consistently shows that companies integrating ESG criteria into executive incentives outperform their peers over longer time horizons. For Nvidia and AMD — whose growth trajectories would support absorbing the complexity of integrating ESG metrics — this represents a near-term opportunity to evolve compensation design beyond pure financial performance. For Intel, a new leadership chapter offers the chance to embed stakeholder-oriented incentives from the outset.
Tax Policy, Regulation, and the External Pressure Landscape
Compensation professionals should also be watching the regulatory environment closely. The SEC's Pay vs. Performance (PVP) disclosure rule is increasing transparency around the actual relationship between executive pay and company-reported financial outcomes. Clawback policies create new structural accountability for executive teams when performance is restated or targets are missed. Gelsinger's forfeiture of 368,965 relative TSR PSUs in March 2024, disclosed in Intel's annual report, is a concrete illustration of these mechanisms functioning as designed.
At the legislative level, proposals like the Tax Excessive CEO Pay Act of 2024 — which would have increased corporate tax rates for companies with CEO-to-worker pay ratios exceeding 50:1 — signal growing political attention to executive pay inequality. Portland, Oregon has already implemented a surtax on companies with ratios exceeding 100:1. Compensation committees designing executive packages today should model the financial and reputational exposure of alternative pay structures against this evolving backdrop.
What This Means for Compensation Professionals
Three practical takeaways emerge from this analysis for HR and compensation practitioners designing or reviewing executive pay programs in capital-intensive, high-growth sectors.
The mix of equity vehicles is as important as total package size. The Nvidia-AMD-Intel comparison makes clear that pure PSU structures work when performance conditions are well-designed and stretch targets are genuinely stretching. RSU components provide retention value, but they can dilute performance pressure when a company most needs it. Be intentional about the ratio.
Performance metrics must map to the actual strategic problem. Intel's metrics were not wrong in isolation — revenue and cash flow matter for a turnaround. But the weighting and surrounding context did not create sufficient urgency. When designing turnaround compensation, consider whether the incentive architecture creates asymmetric urgency proportionate to the challenge being faced.
Stakeholder alignment is not a values statement — it is an incentive design challenge. Integrating even a modest weighting of non-financial metrics (employee retention post-M&A, carbon-reduction milestones, diversity in the leadership pipeline) into long-term incentive plans is operationally feasible and strategically defensible.
The chip industry's CEO pay story between 2021 and 2024 is ultimately a story about alignment — where it worked brilliantly, where it failed in plain sight, and where the next generation of reward design must go.
Gagne, O., Wang, Z., Yu, H., & Allen, D. R. (2025). Comparative Case Research on CEO Compensation: NVIDIA vs. AMD vs. INTEL. Paper presented at the 62nd Eastern Academy of Management, Baltimore, MD, May 2025.