Filing: Vancouver, WA-based Powerlattice, which makes a power delivery chiplet to reduce AI compute power needs by 50%+, raised $86M, after raising $31M in 2025

AI hardware now faces a significant power bottleneck that could slow its expansion.

Powerlattice, a Vancouver, Washington-based startup, has raised $86 million in new funding, according to an SEC filing. This latest investment follows a $31 million round raised in November 2025. The company specializes in developing power delivery chiplets designed to significantly reduce the energy consumption of AI processors. Powerlattice's technology aims to address the growing power and heat limitations faced by data centers due to the intense demands of AI computation. The founders, semiconductor veterans from companies like Qualcomm and Intel, established Powerlattice in 2023. The company recently announced a monolithic integrated voltage regulator chiplet, capable of powering next-generation AI accelerators by optimizing power delivery directly within the processor package.

The substantial funding indicates strong investor confidence in Powerlattice's approach to solving the critical power and thermal challenges in AI compute. This success signals a growing market opportunity for specialized hardware solutions that can enhance efficiency, potentially leading to increased investment in similar chiplet technologies. As AI models become more complex, companies are actively seeking ways to overcome the physical limitations of current hardware, making Powerlattice's advancements a key development.

This development directly impacts AI hardware designers and data center operators by offering a tangible solution to escalating power costs and infrastructure constraints. It suggests a future where chiplet integration becomes more prevalent, allowing for customized and efficient power management tailored to specific AI workloads. Watch for further validation of this technology as it moves from samples to broader adoption, potentially influencing the design of future AI accelerators and the architecture of high-performance computing.