If you pay attention to what major technology companies are saying about the demand for AI, you’ll notice a common thread: a lack of computing power. This means that the large language models that power today’s AI products require more data centers for training and inference, and therefore more power. Against this backdrop, energy efficiency has suddenly become a key priority for semiconductor manufacturers.
PowerLattice, a startup founded in 2023 by veteran electrical engineers from Qualcomm, NUVIA, and Intel, claims to have developed a breakthrough approach that reduces the power demand of computer chips by more than 50%. On Monday, the startup emerged from stealth with a $25 million Series A funding round led by Playground Global and Celesta Capital, bringing total funding to $31 million.
“This is a difficult thing. How do you power a device? There are very few teams and very few people who can do that,” said Pat Gelsinger, general partner at Playground Global. “We have assembled what can be described as a power delivery dream team.”
As the former CEO of Intel, Mr. Gelsinger has great authority in the semiconductor industry, and his participation is a strong seal of approval for PowerLattice.
In fact, when the startup’s CEO Dr. Peng Zou and his founding team pitched the idea at Playground’s offices in March, they were so star-struck by Gelsinger’s fame that they asked him to take a selfie with them, Gelsinger told TechCrunch. The admiration proved mutual, as Mr. Gelsinger came away truly impressed with PowerLattice’s technology.
The startup’s technology sounds simple in concept. It is a small power delivery chiplet designed to bring power closer to the processor, greatly minimizing energy loss.
Two years on, PowerLattice has already achieved its first significant milestone. The first batch of chiplets is being produced by TSMC in partnership with an unnamed manufacturer that is testing the startup’s capabilities, Gelsinger said.
tech crunch event
san francisco
|
October 13-15, 2026
Beyond its original customers, the company plans to allow other customers to test its products in the first half of 2026. The trial should prove fruitful, given that PowerLattice’s potential customer set includes major chip makers Nvidia, Broadcom, and AMD, as well as specialized AI chip developers such as Cerberus, Grok, and Playground-backed startups d-Matrix and NextSilicon.
Every semiconductor company has internal teams working on improving energy efficiency, and Gelsinger hopes PowerLattice’s innovative approach will interest them.
“They might say, ‘We’re going to add a certain amount to this approach and a certain amount to a more traditional approach,'” he says. “However, we believe our ability to capture meaningful share will be immediate.”
PowerLattice isn’t the only startup trying to help chipmakers address energy issues. The company will be in the closest race with Empower Semiconductor, a startup that raised $140 million in Series D led by Fidelity Management & Research Company in September.
However, Gelsinger believes PowerLattice’s 50% energy efficiency improvement is an “extraordinary” result and expects the company to raise even larger funds soon to finance production.
“This idea is bold and the benefits are huge. Others will say, ‘That’s a great idea. I’ll give it a try,'” Gelsinger said.
