MatX raises $500M to challenge Nvidia in AI chip race
What’s the deal? MatX, an AI chip startup founded by two former Google semiconductor engineers, has raised more than $500 million to build hardware aimed at competing with Nvidia.
The round was led by Jane StreetDealroom has a profile for this one. Try Dealroom → and Situational Awareness, the investment firm founded by former OpenAI researcher Leopold AschenbrennerDealroom has a profile for this one. Try Dealroom →. Backers also include Marvell TechnologyDealroom has a profile for this one. Try Dealroom →, NFDGDealroom has a profile for this one. Try Dealroom →, Spark Capital, and Stripe co-founders PatrickDealroom has a profile for this one. Try Dealroom → and John CollisonDealroom has a profile for this one. Try Dealroom →.
MatX was founded by Reiner PopeDealroom has a profile for this one. Try Dealroom → and Mike GunterDealroom has a profile for this one. Try Dealroom →, who left Google in 2022 to design a new chip focused on running large language models. The startup plans to complete its final chip design this year and begin shipping in 2027, working with TSMC for manufacturing.
Why now? Demand for AI computing power continues to surge as companies train and deploy increasingly large models.
Top AI labs such as OpenAI and AnthropicDealroom has a profile for this one. Try Dealroom → are relying on multiple chipmakers and cloud providers, creating openings for challengers to Nvidia’s dominance.
MatX aims to blend two approaches in one product: high-bandwidth memory for training and fast static random access memory for inference. The company says this combination could deliver stronger overall performance.
The new capital will help secure manufacturing capacity and key components ahead of production.
What could go wrong? Competing with Nvidia is a steep challenge.
Chip development is expensive and technically complex, and manufacturing scale-up can expose supply chain constraints, especially in memory components.
MatX must correctly anticipate how AI models evolve and tailor its hardware accordingly. Performance claims will need validation once chips reach customers.
The signal: A $500 million raise for a pre-shipment AI chip startup underscores how investors remain eager to back alternatives to Nvidia.
Capital markets are still rewarding hardware plays tied to AI infrastructure, particularly those targeting leading AI labs.
The funding also highlights a broader shift: AI developers are diversifying suppliers, reducing reliance on a single chipmaker. For startups, that diversification creates rare entry points in a market long dominated by one player.
Sources:
TechCrunch
ITP
Bloomberg
Homebrew
GuruFocus
Intellectia.ai
Image source:
SiliconANGLE
