Palantir Technologies (NASDAQ: PLTR) CEO Alex Karp has identified a narrow group of companies he believes are the definitive winners of the artificial intelligence infrastructure build-out.
Karp placed Palantir alongside Nvidia (NASDAQ: NVDA), Micron Technology (NASDAQ: MU), and SK Hynix (OTCMKTS: SKHY) as the only businesses that truly matter in this new technological era.
At first glance, the grouping appears mismatched, given that each company occupies a distinctly different position across the AI technology stack.
Nvidia supplies the GPUs that train and run AI models, while Micron and SK Hynix dominate the specialized memory market that those same systems depend on.
Palantir sits further downstream, providing the software layer that transforms raw, siloed data into actionable intelligence for corporations and government agencies.
Despite their operational differences, a single financial benchmark ties all four businesses together and reveals their shared trajectory of rapid growth and expanding profitability.
That benchmark is the Rule of 40, a metric that combines a company’s revenue growth rate with its operating profit margin to measure simultaneous financial performance on both fronts.
Palantir’s own Rule of 40 score illustrates just how dramatically the AI era has changed its business, climbing from 83% in the first quarter of 2025 to 145% one year later.
That improvement is driven by two reinforcing forces: accelerating revenue as more commercial customers adopt Palantir’s Artificial Intelligence Platform, and expanding operating margins as fixed costs are spread across a larger revenue base.
New customers therefore add incremental revenue with relatively little additional cost, creating a virtuous cycle that the Rule of 40 is specifically designed to capture.
Nvidia benefits from an analogous dynamic, as its GPUs have become the default engines for AI training and inference inside hyperscale data centers globally, driving explosive revenue growth.
Micron and SK Hynix enjoy parallel tailwinds, since AI workloads require ever-larger volumes of specialized high-bandwidth memory, and insufficient memory causes even the most capable GPU clusters to hit latency bottlenecks.
Each of the four companies required significant up-front capital investment in research, chip fabrication, or software development before reaching their current scale.
Once those foundational investments are in place, however, new revenue flows through the business at high incremental margins, compounding operating leverage over time.
Nvidia, for example, can sell additional GPUs without proportionally increasing its core chip design expenses, allowing profitability to grow faster than its already-rapid top-line expansion.
Karp’s framework suggests that the AI infrastructure cycle is not rewarding all participants equally, but is instead concentrating outsized financial returns among a small number of companies with defensible positions and powerful operating leverage.
