Zain Asgar, a Stanford adjunct professor and founder with a successful exit, has secured an $80 million Series A investment for his new venture. The funding round was spearheaded by Menlo Ventures. His company, Gimlet Labs, has developed what it describes as the inaugural "multi-silicon inference cloud." This software platform enables an AI workload to be executed concurrently across various hardware types.

It can distribute the tasks of an AI application across standard CPUs, specialized AI GPUs, and high-memory systems. As Menlo's lead investor Tim Tully explained in a funding announcement, a single AI agent often involves multiple sequential steps, each with distinct hardware demands: "Inference is compute-bound; decode is memory-bound; and tool calls are network-bound." No single chip currently handles all these needs perfectly. However, with new hardware emerging and older GPUs being repurposed, Tully notes that "the multi-silicon fleet is ready - it’s just missing the software layer to make it work." He believes Gimlet Labs provides precisely that missing layer.

McKinsey projects that if the prevailing trend of simply deploying more computing power continues, data center expenditures could reach nearly $7 trillion by 2030. Asgar points out that applications typically utilize existing deployed hardware only "somewhere between 15 to 30 percent" of the time. "Another way to think about this: you’re wasting hundreds of billions of dollars because you’re just leaving idle resources," he stated. "Our goal was basically to try to figure out how you can get AI workloads to be 10x more efficient than ever, today."

To achieve this, Asgar and his co-founders—Michelle Nguyen, Omid Azizi, and Natalie Serrino—built orchestration software. This technology divides complex agentic workloads so they can be spread simultaneously across diverse hardware. Gimlet Labs asserts its solution reliably accelerates AI inference by 3x to 10x without increasing cost or power consumption. The company says it can even partition the underlying AI model itself, running different segments across the most suitable chip architectures.

Gimlet has already established partnerships with major chip manufacturers, including NVIDIA, AMD, Intel, ARM, Cerebras, and d-Matrix. Its product, offered either as software or via an API to its Gimlet Cloud, is not aimed at everyday AI developers. Instead, it targets the largest AI model laboratories and data center operators. The company launched publicly in October, reporting eight-figure revenues from the start, meaning at least $10 million. Asgar noted that its customer base has more than doubled in the past four months and now includes a major AI model creator and a very large cloud computing firm, though he did not disclose their names.

The founding team previously collaborated at Pixie, a startup that developed an open-source observability tool for Kubernetes. Pixie was acquired by New Relic in 2020, just two months after its launch with a $9 million Series A led by Benchmark. Pixie's technology is now part of the open-source organization that manages Kubernetes.

After Asgar happened to meet Tim Tully about a year ago and also received angel investments from Stanford professors, venture capitalists began reaching out. Following Gimlet's launch, a term sheet was presented. When investors learned Asgar was evaluating offers, "we got a pretty big swarm of funding," he said, and the round quickly became oversubscribed. Including an earlier seed round, the startup has now raised a total of $92 million. Angel investors include Sequoia's Bill Coughran, Stanford Professor Nick McKeown, former VMware CEO Raghu Raghuram, and Intel CEO Lip-Bu Tan. The company currently has a team of 30 employees. Other investors in the round include Factory, which led the seed round, along with Eclipse Ventures, Prosperity7, and Triatomic.