GPU startup Oxmiq has emerged from stealth with plans to launch licensable GPU software and hardware IP following a $20 million seed raise.

Founded two years ago by Raja Koduri, a GPU architect who has previously worked at Apple, AMD, and Intel, Oxmiq’s hardware and software targets AI, graphics, and multimodal workloads.

The startup’s hardware offering consists of a RISC-V-based GPU IP core, called OxCore, that supports near-memory and in-memory compute capabilities, and a chiplet-based system-on-chip (SoC) builder, dubbed OxQuilt, that allows customers to configure optimal compute, memory, and interconnect ratios for their needs.

Meanwhile, Oxmiq’s software offering supports deployment across both Oxmiq IP-based silicon and third-party GPU and AI accelerator platforms. This includes OXCapsule, which abstracts hardware complexity to “provide frictionless deployment across diverse computing platforms,” and OXPython, which allows Python-based CUDA applications to run unmodified on non-Nvidia hardware.

OXPython will first launch on Tenstorrent’s AI platform later this year, with Oxmiq saying that multiple vendor integrations are currently in progress.

“We’re excited to partner with Oxmiq on their OXPython software stack,” said Jim Keller, CEO of Tenstorrent. “OXPython’s ability to bring Python workloads for CUDA to AI platforms like Wormhole and Blackhole is great for developer portability and ecosystem expansion. It aligns with our goal of letting developers open and own their entire AI stack.”