- Odinn Infinity Cube combines multiple Omnia supercomputers in a single glass case
- Memory capacity reaches 86TB of registered DDR5 ECC RAM
- NVMe storage in the cube amounts to a whopping 27.5PB
Odinn, a California-based startup, has introduced Infinity Cube as an attempt to compress data center-class computing into a visually contained structure.
At CES 2026, the company introduced Odinn Omnia, a mobile AI supercomputer, although a system of that scale alone would face clear performance limitations, and this is where the Cube comes into play.
The Infinity Cube is a 14ft x 14ft AI cluster capable of assembling multiple Omnia AI supercomputers into a single glass enclosure.
Scaling AI with modular clustering
This device emphasizes extreme component density rather than incremental efficiency improvements.
According to Odinn, a fully customizable core specification allows the Cube to scale up to 56 AMD EPYC 9845 processors, totaling 8,960 CPU cores.
Its GPU capacity extends to 224 Nvidia HGX B200 units, combined with 43TB of combined VRAM.
For storage, the device supports up to 86TB of DDR5 ECC registered RAM, while the NVMe storage capacity reaches 27.5PB.
These figures imply substantial internal interconnection and power distribution demands that the company has not publicly detailed.
The device uses liquid cooling and each Omnia unit manages its own thermal requirements without shared external infrastructure.
This design avoids reliance on raised floors or centralized cooling plants, at least in theory.
Infinity Cube relies on a proprietary software layer called NeuroEdge to coordinate workloads across the cluster.
The software integrates with Nvidia’s AI software ecosystem and common frameworks, handling programming and deployment automatically.
This abstraction is intended to reduce the need for manual tuning, although it also creates operational dependency on the maturity of the Odinn software.
Institutions that already rely on cloud infrastructure for AI workloads may wonder if on-premises orchestration simplifies management in real-world conditions.
The company says Infinity Cube suits organizations with strict privacy, security or latency requirements that discourage reliance on the cloud.
Placing infrastructure closer to workloads can reduce network delays, but also shifts responsibility for uptime, maintenance, and lifecycle management to the owner.
The idea of ​​presenting data center hardware within compact glass enclosures can be aesthetically appealing.
However, practical trade-offs between density, accessibility and resilience remain unresolved without evidence of real-world implementation.
TechRadar will cover this year’s edition extensively CESand will bring you all the important announcements as they happen. Go to our CES 2026 News page for the latest stories and our hands-on verdicts on everything from wireless TVs and foldable screens to new phones, laptops, smart home devices and the latest in artificial intelligence. You can also ask us a question about the show on our CES 2026 Live Q&A and we will do our best to answer it.
And don’t forget follow us on tiktok and WhatsApp For the latest from the CES fair!
