Top 10 AI Chipmakers Fueling the Intelligence Revolution – Q2 2025 Rankings

State of the Sector: Q2 2025 Overview

AI chips are no longer just technical components — they’re geopolitical assets. Q2 saw a continuation of the compute arms race, with hyperscalers designing their own silicon, sovereign nations investing in domestic fabs, and the market rewarding those with both scale and agility. NVIDIA remains dominant, but competition is intensifying across datacenter, consumer, and edge categories. Meanwhile, China’s domestic chipmakers are gaining ground as U.S. restrictions tighten.

Top 10 Public AI Chipmakers — Q2 2025

1. NVIDIA (NVDA, NASDAQ)

Still the undisputed champion of AI compute. The H200 rollout is exceeding expectations, Blackwell is generating serious buzz, and enterprise adoption is deepening across sectors. Despite stretched valuations, investors remain confident in NVIDIA’s dual strength in datacenter and developer ecosystem.

2. AMD (AMD, NASDAQ)

AMD is closing the gap — not overtaking, but outmaneuvering in smart ways. The new Ryzen AI Pro 300 series has landed strong, and its EPYC chips are winning favor in energy-conscious AI workloads. AMD’s growth story in AI is becoming too strong to ignore.

3. Broadcom (AVGO, NASDAQ)

While less visible than NVIDIA or AMD, Broadcom is everywhere behind the scenes. Their custom AI ASICs power hyperscaler backends from Meta to Google. It’s not the flashiest chip company — but it’s one of the most deeply embedded.

4. Intel (INTC, NASDAQ)

The comeback story continues. Intel’s Gaudi3 accelerators are gaining limited traction, and they’ve made bold moves to reassert relevance in AI-centric datacenter architecture. Q2 saw positive sentiment shift, but execution still needs to catch up to ambition.

5. Qualcomm (QCOM, NASDAQ)

AI at the edge is Qualcomm’s playground. Their Snapdragon X Elite chip is powering the first real wave of AI PCs, and mobile demand remains steady. If AI on-device becomes real for consumers, Qualcomm’s positioning is enviable.

6. Alphabet (GOOGL, NASDAQ)

Google's not just a software company — its TPU architecture continues to drive performance behind every Gemini query. With v6 on the way and custom silicon appearing in more enterprise workloads, Alphabet is a stealth hardware juggernaut in AI.

7. Amazon (AMZN, NASDAQ)

Trainium and Inferentia aren’t household names — but they don’t need to be. Amazon’s vertical AI chip stack gives AWS pricing power and efficiency advantages. It's not a chip play for the public — it’s a moat for cloud dominance.

8. Marvell Technology (MRVL, NASDAQ)

Often overlooked, Marvell is becoming essential in AI networking. Their chips underpin the high-speed data transfer infrastructure necessary for AI training at scale. Q2 brought bullish analyst upgrades based on hyperscaler momentum.

9. Arm Holdings (ARM, NASDAQ)

Since its IPO, Arm has remained central to the energy-efficient AI movement. With Arm-based designs appearing in everything from smartphones to datacenter AI inference chips, its licensing model is quietly dominating at scale.

10. Taiwan Semiconductor Manufacturing Company (TSM, NYSE)

TSMC may not “make” AI chips in the brand sense, but without them, no one else does either. Their cutting-edge fabs are where NVIDIA, AMD, and Apple silicon come to life. They’re the manufacturing heartbeat of the entire AI race.

Private Companies to Watch

While public investors can't directly tap into these firms yet, they are shaping the next wave of silicon breakthroughs:

Cerebras Systems: Their CS-3 wafer-scale engine pushes the boundaries of AI training speeds. Already used by Mistral and Perplexity AI, it’s a giant in every sense.
Tenstorrent: Led by chip legend Jim Keller, Tenstorrent is designing general-purpose AI compute cores with ambitions across datacenter and edge.
Groq: Known for ultra-low latency processors ideal for inference. With recent announcements around GroqChat, they're making waves in LLM speed benchmarking.
Mythic AI: Specializing in analog AI processing for edge devices, Mythic offers an intriguing, energy-efficient take on local inference.
Etched AI: An upstart focused on fixed-function LLM chips. If specialization wins over generalization, Etched could leapfrog legacy designs.

Neural Capital Insight

If Q1 was about excitement, Q2 is about consolidation. The players on this list aren't just chasing AI — they're defining how it's built and where it runs. Custom silicon is the new competitive edge, and companies owning their own chip destiny are pulling ahead. As more nations and corporations seek compute independence, chipmakers are gaining not just market value — but geopolitical weight.

Get Educated - Get Smart - Be Ready

Track every quarterly shift in AI leadership with NeuralCapital.ai. Check back July 1st for the Q3 2025 Power Rankings across all 20 AI sectors.