Row-Scale On-Premises Cloud Infrastructure from Oxide Computer

Row-Scale On-Premises Cloud Infrastructure from Oxide Computer

The News: Oxide presented at Cloud Field Day 20, showcasing a rack-scale architecture to deliver on-premises cloud infrastructure that is fully integrated and power efficient. Watch all the Oxide presentations at Tech Field Day.

Row-Scale On-Premises Cloud Infrastructure from Oxide Computer

Analyst Take: A rack full of high-density on-premises cloud infrastructure might be an exciting prospect. For large enterprises, multiple rows of these racks are essential; this is where Oxide plays. Our visit to the Oxide headquarters during Cloud Field Day 20 showed us that Oxide started with first-principles hardware design for the entire infrastructure. A clean-slate design of rack-scale infrastructure allowed some legacy constraints to be eliminated where they don’t deliver benefits, such as allowing larger, more energy-efficient fans. Oxide uses the rack-scale architecture as a deployment unit. How many racks of cloud did you want?

Infrastructure as a Service

The Oxide rack delivers the fundamental food groups of on-premises cloud infrastructure IaaS: compute, storage, and networking. Each Oxide rack provides all the resources with 2048 Cores, 32,000 GB of RAM, nearly a petabyte of NVMe SSD storage, and 12 Tbps of network throughput. The entire rack is designed as an integrated unit for power efficiency. A pair of beefy power supplies take in three-phase AC power and deliver redundant DC power to everything in the rack. A smaller number of larger AC power supplies are far more efficient than having power supplies in every server. The DC bus-bar distribution also requires fewer cables in the back of the rack, freeing air flow and reducing the power needed for cooling. The 100 GB ethernet to the servers is carried over a backplane, again simplifying cabling and improving airflow. I cannot emphasize enough how each rack is an integrated system with many incremental improvements over conventional racks of servers.

Everything Open All the Time

One of Oxide’s unusual aspects is its core values of openness and humility. All the software that runs the Oxide rack is open source on GitHub for customers and anyone else to view and improve. There is a culture of building robust, supportable, debuggable hardware and providing everything to enable customers to operate their Oxide platform as on-premises cloud infrastructure. The first-principles redesign is shown in the removal of legacy technologies. The management processor isn’t a CPU running a Linux operating system; it’s a much simpler microcontroller with a tiny firmware. Much of the simplification is possible because everything in the rack is pre-defined by Oxide. There are no empty PCIe slots for adding different NICs or GPUs. The servers do not have a BIOS; the operating system initializes the standard hardware. Once you roll the rack in place, you are unlikely to make any hardware changes; replace anything that fails.

Who Is Oxide for?

The customer who needs the Oxide solution is looking for cloud-scale infrastructure in their own data center. They might be a global 200 organization operating multiple megawatt data centers and wanting to spend less time managing the servers and must limit their power usage. They might be regional cloud providers, where the Oxide platform is the basis for their cloud services. Another common factor is the desire to enable self-service consumption of resources by internal staff or tenants—cloud-like consumption with on-premises cloud infrastructure.

Disclosure: The Futurum Group is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.

Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of The Futurum Group as a whole.

Other Insights from The Futurum Group:

Liqid: The Right Place, Right Time for Composable AI Stack

The Evolving Landscape of Private Cloud Infrastructure

Introducing TruScale Backup with Veeam: Lenovo’s New On-Premises Cloud Solution

Author Information

Alastair has made a twenty-year career out of helping people understand complex IT infrastructure and how to build solutions that fulfil business needs. Much of his career has included teaching official training courses for vendors, including HPE, VMware, and AWS. Alastair has written hundreds of analyst articles and papers exploring products and topics around on-premises infrastructure and virtualization and getting the most out of public cloud and hybrid infrastructure. Alastair has also been involved in community-driven, practitioner-led education through the vBrownBag podcast and the vBrownBag TechTalks.


Latest Insights:

The Six Five team discusses Sequoia/A16Z/Goldman rain on the AI parade.
The Six Five team discusses Oracle & Palantir Foundry & AI Platform.
The Six Five team discusses AWS Summit New York 2024.