Funding Round and Leadership

Gimlet Labs, a startup focused on AI inference optimization, closed a Series A financing round of $80 million. The round was led by Menlo Ventures, with participation from a group of angels and other venture firms. Founder and Stanford adjunct professor Zain Asgar, together with co‑founders Michelle Nguyen, Omid Azizi, and Natalie Serrino, will use the capital to expand the company’s multi‑silicon inference platform.

Technology Overview

The core of Gimlet Labs’ offering is a software layer it calls the “multi‑silicon inference cloud.” This layer orchestrates AI workloads across diverse hardware types—traditional CPUs, AI‑tuned GPUs, and high‑memory systems—allowing a single AI task to be split into steps that each run on the hardware best suited for that step. According to Menlo Ventures’ Tim Tully, inference is compute‑bound, decoding is memory‑bound, and tool calls are network‑bound, and the platform dynamically matches each requirement to the appropriate silicon.

Performance Claims

Gimlet Labs asserts that its solution can accelerate AI inference by three to ten times without increasing cost or power consumption. The platform can also partition the underlying model so that different portions execute on different architectures, selecting the optimal chip for each segment. The company reports that existing data‑center hardware is currently utilized only between 15 % and 30 % of the time, leading to wasted resources worth “hundreds of billions of dollars.” By better leveraging idle capacity, the startup aims to make AI workloads up to ten times more efficient.

Market Position and Partnerships

The product is delivered either as software or via an API to Gimlet’s own cloud service, targeting large AI model laboratories and major data‑center operators rather than everyday AI developers. Gimlet Labs has already formed partnerships with leading chip manufacturers including NVIDIA, AMD, Intel, ARM, Cerebras, and d‑Matrix, bolstering its multi‑hardware approach.

Company Background and Growth

Before Gimlet Labs, the founding team worked together at Pixie, an open‑source observability tool for Kubernetes that was acquired by New Relic in 2020. Gimlet Labs launched publicly in October, reporting eight‑figure revenue out of the gate and a customer base that has more than doubled in four months, including a major model maker and a large cloud computing company (names undisclosed). The company now employs 30 people.

Investor Landscape

In addition to Menlo Ventures, the round attracted investors such as Sequoia’s Bill Coughran, Stanford professor Nick McKeown, former VMware CEO Raghu Raghuram, Intel CEO Lip‑Bu Tan, Factory (seed lead), Eclipse Ventures, Prosperity7, and Triatomic. The total capital raised by Gimlet Labs to date is $92 million.

Used: News Factory APP - news discovery and automation - ChatGPT for Business