The San Jose startup wants to cut custom AI silicon development time and cost in half with modular chiplets.
Building a custom AI chip from scratch is roughly as fun as it sounds: expensive, slow, and full of opportunities for things to go sideways. TYLSemi just raised $43 million betting it can make the process a lot less painful.
The San Jose-based semiconductor startup announced the oversubscribed early-stage round on July 14, with Matter Venture Partners leading the deal. Viola Ventures, GHOVC, Egis Technology, and a handful of strategic investors focused on semiconductor and AI infrastructure also participated.
What TYLSemi actually does #
The company is developing a full-stack chiplet platform, building standardized, reusable chip components that snap together like building blocks. Instead of designing every piece of an AI chip from scratch, companies can grab pre-built modules off TYLSemi’s shelf and focus their engineering effort on the parts that actually differentiate their product.
The product lineup includes three core offerings. TYL.IO handles connectivity between chiplets. TYL.Power manages intelligent power delivery. And TYL.Forge ties everything together as the full-stack custom silicon production platform.
These chiplets are built on UCIe and other industry standards, designed to play nicely with existing chip architectures and advanced packaging technologies. TYLSemi claims its approach can slash custom AI silicon development time and cost by up to 50%.
The people behind it #
The company was co-founded by Mohit Gupta, who serves as CEO, and Sunil Bhardwaj. Both bring experience from Alphawave (which was acquired by Qualcomm), SiFive, Cadence Design Systems, and Rambus.
TYLSemi has lined up a collaboration with TSMC to produce samples of its TYL.IO and TYL.Power chiplets. Those first samples are expected to arrive in 2027.
Why this matters for the broader AI hardware race #
The AI accelerator market is projected to reach $604 billion by 2033, with custom silicon XPUs expected to be the fastest-growing segment. Every hyperscaler, every major cloud provider, and increasingly every large enterprise wants chips tailored to their specific AI workloads rather than relying solely on general-purpose GPUs.
TYLSemi is positioning itself as the first dedicated provider of a comprehensive chiplet portfolio, lowering the barrier to entry for organizations that want custom AI silicon but lack the resources to build everything in-house.
The $43 million round was oversubscribed, suggesting demand from investors exceeded the amount TYLSemi was initially seeking. TYLSemi’s differentiation rests on being purpose-built for AI workloads and offering a dedicated, standards-based chiplet portfolio.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our