Oxmiq Raises $35M for GPU IP, Expands Focus to Data Center Design

2026-07-01 · Source: Big Data & AI News - EE Times · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure · Depth: Advanced, medium

Summary

Oxmiq Labs, the GPU hardware and software startup founded by Raja Koduri, has secured \$35 million in Series A funding, bringing its total capital raised to \$60 million. This investment will fuel the company's expansion into custom data center-scale hardware and software, including bespoke silicon. Oxmiq aims to build a comprehensive technology stack, from energy generation to AI factory infrastructure design, featuring its OxCore GPU IP, "Electron to token machines" (custom appliances), and AI data center pod designs. The company is also developing OxFactory, a data center-scale orchestration software, alongside its existing OxCapsule for edge hardware and OxPython for bridging Nvidia GPU porting. Oxmiq's strategy involves enabling hyperscalers to save billions by offering custom silicon and direct foundry access, exemplified by its partnership with AM Intelligence Labs for a 2-GW AI deployment in India, with a 1-GW cluster coming online in late 2027. The company, currently with 45 employees, plans to grow to 60-70, emphasizing productivity through experienced personnel and agentic tools.

Key takeaway

For AI Architects and CTOs evaluating future data center builds, Oxmiq's expanded offerings signal a viable path to significantly reduce AI infrastructure costs. You should consider custom silicon and direct supply chain models, potentially saving billions on 100-MW deployments. Explore partners like Oxmiq for bespoke GPU IP and chiplet-based designs. This approach allows optimization for specific throughput, latency, or budget constraints, moving beyond traditional vendor lock-in.

Key insights

Oxmiq is expanding its GPU IP and software stack to offer custom, cost-efficient AI data center designs and bespoke silicon for hyperscalers.

Principles

• Custom silicon and direct supply chain significantly reduce AI infrastructure costs.
• Heterogeneous hardware orchestration is critical beyond CUDA for data center efficiency.
• Small, expert teams with agentic tools can achieve high productivity in complex tech.

Method

Oxmiq designs full AI factory stacks, from GPU IP (OxCore) and custom appliances ("Electron to token machines") to data center pod designs, using OxFactory orchestration and OxQuilt for chiplet customization.

In practice

• Explore custom silicon options to reduce AI data center capex by billions.
• Implement heterogeneous orchestration software for non-Nvidia GPU deployments.
• Utilize chiplet architectures like OxQuilt for workload-specific hardware optimization.

Topics

• GPU IP
• Data Center Design
• AI Infrastructure
• Custom Silicon
• Oxmiq Labs
• Chiplet Architecture
• AI Orchestration

Best for: VP of Engineering/Data, Director of AI/ML, Investor, AI Hardware Engineer, AI Architect, CTO

Related on AIssential

• Why Everyone Is Moving Away from NVIDIA — Hyperscalers are vertically integrating AI infrastructure, developing custom silicon and energy solutions to overcome GPU bottlenecks.
• OpenAI’s Jalapeño Will Be Spicy, But the Real Sizzle Is Its Chip Design AI — OpenAI's Jalapeño chip and AI-driven design process signal a strategic shift in compute infrastructure and development.
• Exclusive: Allen Wu on Disrupting $100M Cost of Building Custom AI Chips — Heterogeneous custom AI chip development can be made affordable and accessible using an open, modular RISC-V baseline.
• After Nvidia’s $20B not-acqui-hire, AI chip startup Groq reportedly raising $650M — Groq is pivoting to an inference neocloud model, securing significant funding after a strategic Nvidia deal.
• Meta ramps up AI push with expanded Broadcom agreement - Latest news from Azerbaijan — Meta and Broadcom are deepening their custom AI chip partnership to scale Meta's AI computing infrastructure.
• ChatGPT Designs Cancer Vaccine — Semiconductor manufacturing, particularly EUV lithography, is the ultimate bottleneck for AI compute scaling, not power or data centers.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Big Data & AI News - EE Times.