Lonjakan Pendapatan AI: Tiga Pemimpin Infrastruktur Siap Mendominasi Gelombang Berikutnya

The artificial intelligence sector is experiencing unprecedented growth, and this momentum is clearly reflected in recent AI earnings reports. With companies racing to build out their infrastructure, now is an ideal time to examine which firms are positioned to capitalize on this explosive expansion. Three key players—Nvidia, Broadcom, and Advanced Micro Devices—have emerged as frontrunners in the competitive race to power the AI revolution, each with distinct competitive advantages and market opportunities.

Nvidia’s Commanding Position in AI Earnings Growth

When it comes to controlling the AI chip market, Nvidia stands in a league of its own. The company’s most recent quarter tells the story—a remarkable 63% surge in revenue to $57 billion demonstrates just how dramatically the AI boom is reshaping the company’s financial trajectory. At Nvidia’s scale, this kind of growth is virtually unprecedented, and it underscores the insatiable demand for artificial intelligence infrastructure.

Yet the most striking part of Nvidia’s AI earnings story isn’t even the GPU revenue. Its data center networking division has become a growth engine in its own right, expanding by 162% to reach $8.2 billion in a single quarter. This segment encompasses Nvidia’s NVLink interconnect system, InfiniBand, and Spectrum-X Ethernet products—essentially allowing the company to sell complete end-to-end AI solutions it brands as “AI factories.”

The reason Nvidia maintains such an iron grip on the market goes beyond just making powerful chips. Its proprietary CUDA software platform has become the industry standard for foundational AI code development. Additionally, the NVLink interconnect solution locks customers into Nvidia’s ecosystem by enabling GPUs to function as a unified system, making it impractical to mix competitors’ chips within an AI cluster. This combination of technical moat and infrastructure lock-in makes Nvidia’s AI earnings growth particularly sustainable.

Broadcom’s Custom Chip Opportunity in the AI Infrastructure Race

While Nvidia dominates the headlines, Broadcom is quietly positioning itself for substantial AI earnings growth through a fundamentally different approach. The company possesses a powerful networking portfolio of components that manage data flow and distribute AI workloads across server architectures. This business is growing steadily, but Broadcom’s most significant opportunity lies in a different arena: helping hyperscalers develop their own custom artificial intelligence chips.

These custom chips, known as ASICs (application-specific integrated circuits), represent a compelling alternative to GPUs for certain workloads—particularly inference, which represents an ongoing operational cost rather than a one-time training expense. The economic logic is compelling: ASICs consume substantially less power than GPUs, dramatically improving cost-effectiveness for inference operations. The trade-off is flexibility; since they’re pre-programmed for specific tasks, they can’t adapt if technology shifts or requirements change.

Broadcom’s track record here speaks volumes. The company helped Alphabet create its Tensor Processing Units (TPUs), which became highly influential in the market. This success attracted additional hyperscalers seeking custom chip design assistance. Broadcom disclosed that three customers further along in development represent a market opportunity worth $60-90 billion by fiscal 2027. To underscore the revenue potential, a fourth customer recently placed a $10 billion order for the following year. The company has also secured a major contract with OpenAI, signaling confidence in its infrastructure strategy. These agreements position Broadcom to capture meaningful AI earnings growth over the coming years.

AMD’s GPU Expansion Strategy in a Growing Market

Advanced Micro Devices is approaching the AI infrastructure opportunity through yet another lens. The company leads in data center central processing units (CPUs)—the “brains” that work alongside GPUs. While the data center market is expanding rapidly, AMD’s most significant growth vector lies in capturing greater share of the GPU market as inference workloads become increasingly important to operators’ economics.

AMD has successfully carved out a niche in this space, and recent developments suggest it’s positioned to expand its footprint. The company has entered into a compute partnership with OpenAI, demonstrating its technological viability. More significantly, Microsoft is developing tools to convert Nvidia’s proprietary CUDA code to AMD’s ROCm software platform specifically for inference applications—a move that could substantially lower the switching costs for adopting AMD solutions.

AMD has established ambitious long-term targets for market expansion. While the company may only capture a fraction of the massive GPU accelerator market, even modest share gains in this enormous and rapidly expanding segment would generate substantial AI earnings growth. The company’s strategic positioning and technological partnerships suggest it has real potential to capitalize on the ongoing infrastructure buildout.

The Path Forward for AI Infrastructure Investors

The common thread connecting these three companies is straightforward: artificial intelligence infrastructure spending shows no signs of slowing. Each company has identified distinct advantages in this competitive landscape. Nvidia’s technical moat and system lock-in create a defensible position at the premium end of the market. Broadcom’s ASIC strategy opens a multi-billion-dollar opportunity with major hyperscalers committing substantial capital. AMD’s push into GPUs, supported by Microsoft’s CUDA-to-ROCm conversion tools, represents a genuine threat to Nvidia’s market share in the inference segment.

For investors focused on AI earnings growth, understanding these differentiated strategies is essential. The companies best positioned to capture this expanding market opportunity are those that have already demonstrated execution capability and secured major customer commitments. Each of these three infrastructure leaders has accomplished exactly that, making them worthy of consideration as the AI infrastructure buildout accelerates.