Company intelligence
Oxmiq Labs
Re-Architecting the GPU Stack: From Atoms to Agents™
About Oxmiq Labs
Founded by Raja Koduri, OXMIQ™ is rearchitecting the GPU full stack from Atoms to Agents to meet the demands of next-generation gaming, graphics and multimodal AI. The company develops licensable GPU software and hardware IP that balances flexibility and performance by integrating breakthrough technologies—including nano agents in silicon based on RISC-V, near- and in-memory computing, advanced light transport, and other innovations. OXMIQ’s architecture is designed to scale seamlessly from physical AI devices all the way to Data Center scale. See oxmiq.ai for more info.
Verified activity
Signals from Oxmiq Labs
19 published signals
Products & Services
Oxmiq Labs announced that Nebius and Palantir are offering a bundled stack including infrastructure and operational intelligence tools designed to meet data residency and security requirements.
Reported by Mark Hirsch
Partnerships
Oxmiq Labs announced that Palantir brings the software layer.
Reported by Mark Hirsch
Partnerships
Oxmiq Labs announced that Nebius brings NVIDIA-based GPU clusters across the U.S. and Europe.
Reported by Mark Hirsch
Customers & Market
Oxmiq Labs notes that Anthropic has committed contracts spanning Akamai, AWS, CoreWeave, Fluidstack, Google Cloud, Lambda, Volta, and Microsoft, covering 14.8 gigawatts of capacity over the next decade.
Reported by Mark Hirsch
Research & Knowledge
Oxmiq Labs noted that DeepSeek's R1 model, released in January 2025, was trained on NVIDIA H800 chips and demonstrated competitive performance.
Reported by Mark Hirsch
Products & Services
Oxmiq Labs announced DeepSeek deployment of 160,000 chips in Inner Mongolia for inference workloads targeting a pilot project.
Reported by Mark Hirsch
Research & Knowledge
Oxmiq Labs announced that the goal is 1.6 terabit DR8 optical links in 2026, scaling to 3.2T by end of 2027.
Reported by Mark Hirsch
Research & Knowledge
Oxmiq Labs published a QYResearch report naming Google, Meta, and NVIDIA as key players in the robotic agent segment.
Reported by Mark Hirsch
Products & Services
Oxmiq Labs notes that RTX Spark laptops are shipping this fall.
Reported by Mark Hirsch
Products & Services
Oxmiq Labs notes that Microsoft unveiled RTX Spark at Computex in late May and detailed the Windows integration at Build 2026 in early June.
Reported by Mark Hirsch
Research & Knowledge
Oxmiq Labs reported that HydraFusion delivered a 67% cost reduction and a 4.9 percentage point improvement in verified task quality on TerminalBench 2.1 versus Claude Opus 5.
Reported by Mark Hirsch
Customers & Market
Oxmiq Labs announced that Nscale is building AI data centers and selling GPU capacity at scale with Figure for up to 100,000 GPUs on NVIDIA's Vera Rubin platform, with deployment starting in the second half of 2027.
Reported by Mark Hirsch
Research & Knowledge
Oxmiq Labs published a report analyzing the cost advantage of Chinese AI models delivering 90-95% of the capability of frontier U.S. systems at under 10% of the training cost.
Reported by Mark Hirsch
Customers & Market
Oxmiq Labs operates in 35+ markets.
Reported by Mark Hirsch
Capital & Finance
Oxmiq Labs reported that the company currently has 650 people and expects to reach roughly 1,000 employees per reporting from WSJ and Bloomberg this week.
Reported by Mark Hirsch
Capital & Finance
Oxmiq Labs raised $550 million at this valuation, with total capital raised now over $800 million since founding in early 2025.
Reported by Mark Hirsch
Capital & Finance
Oxmiq Labs is at roughly $70 million in ARR and expects to hit $100 million by year-end.
Reported by Mark Hirsch
Products & Services
Oxmiq Labs announced the SN50, a chip combining 432 MB of on-chip SRAM, 64 GB of HBM2E on-package, and up to 512 GB of DDR5 on the accelerator card.
Reported by Mark Hirsch
Partnerships
Oxmiq Labs announced an expanded collaboration between Tencent Cloud and Elastic, integrating Elastic's search and observability platform deeper into Tencent Cloud's AI services.
Reported by Mark Hirsch