Key Takeaways
- • Gimlet Labs closed a $300 million Series B funding round led by Andreessen Horowitz at a $3 billion valuation, bringing total funding to $392 million.
- • The round attracted 17 institutional investors including strategic hardware partners Arm Holdings and M12 (Microsoft's venture fund), alongside leading venture firms Sapphire Ventures, Menlo Ventures, and Tiger Global Management.
- • The company has secured billions of dollars in contracted revenue for Gimlet Cloud and plans to deploy the capital toward scaling multi-silicon cloud capacity, securing gigawatts of high-density data center infrastructure, and expanding serverless/managed cloud deployments.
- • Gimlet Labs operates a disaggregated AI inference platform that splits complex neural network workloads across heterogeneous hardware accelerators (GPUs, near-memory compute, dataflow architectures, CPUs) to optimize for ultra-low latency and power efficiency.
- • Co-founded by CEO Zain Asgar, Michelle Nguyen, Omid Azizi, Natalie Serrino, and James Bartlett, the San Francisco-based company is positioning itself as critical infrastructure for enterprise AI deployment across multi-chip environments.
Summary
Gimlet Labs, a San Francisco-based AI infrastructure provider, has raised $300 million in Series B funding at a $3 billion post-money valuation, according to primary filings and publisher disclosures. The round was led by Andreessen Horowitz (a16z) and includes participation from 16 co-investors spanning venture capital, strategic corporate, and financial institutions. Notable backers include Arm Holdings and M12 (Microsoft's venture arm)—both strategic hardware partners—alongside established venture firms Sapphire Ventures, Menlo Ventures, Tiger Global Management, and Hudson River Trading. This brings Gimlet Labs' total funding to $392 million.
The company develops a multi-silicon inference cloud platform designed to orchestrate AI workloads across heterogeneous hardware architectures. Gimlet's technology employs advanced disaggregation techniques to break complex neural networks into granular processing phases, distributing compute-intensive "prefill" cycles separately from memory-bandwidth-constrained "decode" cycles, as well as attention-FFN splits and speculative-decode operations. The unified software stack programmatically routes each workload segment to optimal mixed-hardware clusters, including GPUs, near-memory compute architectures, dataflow processors, and CPUs, enabling ultra-low latency execution with maximum power efficiency.
Gimlet Labs has already secured billions of dollars in contracted revenue commitments for its Gimlet Cloud platform, indicating substantial customer traction prior to this funding round. The capital will be deployed to expand multi-silicon cloud capacity, secure gigawatts of high-density data center pipeline infrastructure, expand serverless and managed cloud deployments, and scale the organization across systems architecture, data center operations, distributed networks, and hardware compiler teams.
The five co-founders—CEO Zain Asgar, Michelle Nguyen, Omid Azizi, Natalie Serrino, and James Bartlett—are positioning the company as foundational infrastructure for enterprise AI inference in an era of proliferating chip architectures. The participation of both Arm Holdings and Microsoft's venture fund signals institutional recognition of Gimlet's potential to become a critical abstraction layer in the AI compute stack, where hardware fragmentation and power constraints are accelerating demand for software-defined optimization solutions.
The company develops a multi-silicon inference cloud platform designed to orchestrate AI workloads across heterogeneous hardware architectures. Gimlet's technology employs advanced disaggregation techniques to break complex neural networks into granular processing phases, distributing compute-intensive "prefill" cycles separately from memory-bandwidth-constrained "decode" cycles, as well as attention-FFN splits and speculative-decode operations. The unified software stack programmatically routes each workload segment to optimal mixed-hardware clusters, including GPUs, near-memory compute architectures, dataflow processors, and CPUs, enabling ultra-low latency execution with maximum power efficiency.
Gimlet Labs has already secured billions of dollars in contracted revenue commitments for its Gimlet Cloud platform, indicating substantial customer traction prior to this funding round. The capital will be deployed to expand multi-silicon cloud capacity, secure gigawatts of high-density data center pipeline infrastructure, expand serverless and managed cloud deployments, and scale the organization across systems architecture, data center operations, distributed networks, and hardware compiler teams.
The five co-founders—CEO Zain Asgar, Michelle Nguyen, Omid Azizi, Natalie Serrino, and James Bartlett—are positioning the company as foundational infrastructure for enterprise AI inference in an era of proliferating chip architectures. The participation of both Arm Holdings and Microsoft's venture fund signals institutional recognition of Gimlet's potential to become a critical abstraction layer in the AI compute stack, where hardware fragmentation and power constraints are accelerating demand for software-defined optimization solutions.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.