Mayfield Adds Chip-Design Specialist to Bet on Mega-Seed Infrastructure Rounds

From Cerebras to Mayfield

Adit Singh has joined Mayfield as an infrastructure partner, the 57-year-old venture firm announced this week. Singh brings a track record of early-stage hardware bets, most notably co-leading the first funding round into Cerebras when the chip startup was still largely unknown. That company went public in May and now trades at a valuation approaching $50 billion, with Singh's former firm holding a roughly 7% stake that made it the third-largest shareholder at IPO.

At Mayfield, Singh will focus on hardware, infrastructure software, cybersecurity, and physical AI. His background as an electrical engineer and chip designer informs his investment approach. "My secret sauce is being able to look at any workload and see how it works from the application all the way down to the transistor," Singh explained in a recent interview. That lens helped him spot potential in Cerebras a decade ago, when few investors understood the company's vision for building the world's largest AI training chip.

Why Mayfield

Singh's path through venture has been winding. After his early success at Foundation Capital, he co-founded Neotribe Ventures in 2017, then spent time at early-stage firm Cota Capital. The move to Mayfield, he says, is driven by a structural shift in how infrastructure deals are priced. Seed rounds for hardware and AI infrastructure startups have ballooned in recent years, often reaching $15 million to $20 million or more. Smaller funds that once competed in the seed category now find themselves priced out.

Mayfield manages $3 billion in assets, giving it the capacity to write seed checks up to $20 million. That scale matters in a market where a promising chip startup or physical AI company can command valuations that would have been reserved for Series A just a few years ago. Singh pointed to Mayfield's semiconductor portfolio as a key draw. The firm has invested in Upscale AI, recently valued at $2 billion, and Lumilens, which raised $700 million at a $5.5 billion valuation.

Mayfield managing partner Navin Chaddha noted that the two have known each other for 15 years and served together on three startup boards, including Upscale AI and Velaura AI, a unicorn AI company. "We have been talking for a long time, and this time stars aligned," Chaddha said.

The Mega-Seed Phenomenon

The rise of mega-seed rounds in infrastructure reflects a broader recalibration in venture capital. Hardware startups, particularly those building custom silicon or robotics platforms, require significant upfront capital for tape-outs, pilot production runs, and engineering teams. At the same time, the potential returns from a breakout chip company have grown exponentially. Cerebras, Groq, and a wave of AI accelerator startups have demonstrated that specialized hardware can command enormous valuations if the technology delivers on performance and power efficiency.

For investors, this creates a dilemma. Traditional seed funds, often managing $100 million to $300 million, struggle to lead rounds that require $15 million or more in initial capital. Larger multi-stage firms, meanwhile, have historically focused on Series A and beyond. Mayfield is positioning itself in the middle: large enough to lead mega-seed deals, but still committed to early-stage investing.

Singh's hire signals that Mayfield intends to compete aggressively in this category. His technical fluency in chip design and systems architecture gives the firm an edge in evaluating startups that may still be in the design phase, with little more than simulations and a founding team to assess.

From Transistor to Application

Singh's engineering background sets him apart in a venture landscape often dominated by generalists. He describes his approach as working backward from the application layer to the transistor level, understanding how software workloads map to hardware architectures. That skill was critical in evaluating Cerebras, which built a wafer-scale chip that defied conventional wisdom about yield, cooling, and interconnects.

The same analytical framework applies to the next generation of infrastructure startups. Physical AI companies, for example, need custom accelerators that can handle sensor fusion, real-time inference, and edge computing constraints. Cybersecurity infrastructure increasingly relies on hardware-based trust anchors and cryptographic co-processors. Singh's ability to assess these technical layers quickly is a strategic asset for Mayfield as it deploys capital into a crowded and complex market.

Mayfield's Semiconductor Bet

Mayfield has quietly assembled one of the stronger semiconductor portfolios in venture capital. The firm's investments span AI accelerators, photonics, and specialized processors for edge and data center workloads. Singh described it as "probably the best in the market right now," a claim that reflects both the firm's recent wins and the broader momentum in custom silicon.

The semiconductor investment cycle is notoriously long and capital-intensive. Startups can spend years in stealth, burning through capital on design, verification, and tape-out before shipping a single chip. But when the technology works, the outcomes can be transformative. Mayfield's willingness to lead large seed rounds and support companies through multiple funding cycles positions it to capture disproportionate returns in a category where patience and technical judgment are as important as capital.

What Comes Next

Singh's move to Mayfield comes at a moment when infrastructure investing is both fashionable and treacherous. The AI boom has driven a surge of interest in hardware, but not every chip startup will find product-market fit. Export controls, geopolitical tensions, and competition from hyperscalers add layers of complexity. Singh will need to navigate these dynamics while building a portfolio that can withstand the inevitable shakeout.

For Mayfield, the bet is clear: the next decade of computing will be defined by specialized hardware, and the firms that can identify and support the best teams early will capture outsized returns. Singh's track record suggests he knows how to spot those opportunities. Whether Mayfield's portfolio delivers another Cerebras-scale outcome remains to be seen, but the firm is positioning itself to compete.