← Back to BLACKWIRE PRISM BUREAU CAPITAL INFLOW Oxide Computer engineers assembling a custom server rack in a cleanroom environment.

Oxide’s open‑hardware server platform, photographed during a pre‑production build in Arizona.

OXIDE COMPUTER SECURES $445M SERIES D, VALUING STARTUP AT $2.5B

*Oxide's latest raise catapults the open‑hardware server maker into unicorn territory. The cash infusion signals deep investor confidence in custom silicon as AI workloads explode.*

By PRISM Bureau - BLACKWIRE  |  October 10, 2026, 06:00 CET  |  Oxide Computer, Series D, custom silicon, AI hardware, cloud infrastructure

Oxide Computer just closed a $445 million Series D, vaulting the San Francisco startup into unicorn status. The raise, led by Andreessen Horowitz and Sequoia Capital, values the company at roughly $2.5 billion—a stark indicator that investors see custom silicon as the next frontier for AI‑driven cloud workloads. Oxide’s pitch is simple: replace generic x86 servers with purpose‑built, open‑hardware racks that run a Rust‑based operating system. The cash will fund a new 64‑core ARM design, expand manufacturing in Arizona, and accelerate hiring across silicon, firmware, and sales. The move comes as hyperscalers double down on in‑house chips, and the market for energy‑efficient AI compute is projected to exceed $12 billion by 2027.

Funding Round Details

Oxide announced a $445 million Series D on Oct. 4, 2024. The round was led by Andreessen Horowitz and Sequoia Capital, with participation from Battery Ventures, DCVC, and existing backers such as General Catalyst. The company disclosed a post‑money valuation of roughly $2.5 billion, a 3.5× jump from its $710 million Series C last year. Oxide’s CEO, Drew Perkins, said the capital will fund a new silicon design cycle and expand its manufacturing footprint in Arizona. The filing listed 42 employees at the time of the raise, up from 28 six months earlier.

Strategic Implications for Cloud Hardware

The infusion directly targets Oxide’s ambition to replace commodity x86 servers with purpose‑built, Rust‑driven platforms. With AI models now demanding teraflops of inference per rack, Oxide’s “Open Compute” blueprint promises lower latency and higher energy efficiency than Amazon’s Graviton or Google’s TPU clusters. The money will finance a 64‑core ARM silicon tape‑out slated for Q2 2025 and a partnership with GlobalFoundries for 5 nm production. If successful, Oxide could capture a slice of the $12 billion cloud‑infrastructure market that currently favors hyperscalers’ in‑house silicon.

"We are building the next generation of cloud infrastructure, not just another server," CEO Drew Perkins said at the announcement.

Risks and Execution Challenges

Scaling custom silicon is fraught with supply‑chain volatility. Oxide must secure wafer capacity at a time when TSMC and Samsung are booked months out. The company also faces a talent bottleneck; only 12 engineers on its silicon team have shipped a tape‑out before. Moreover, integration with major cloud providers requires rigorous certification, a process that can add 12‑18 months to product rollout. Failure to meet the Q2 2025 silicon deadline would erode investor confidence and open the field for rivals like Ampere and Habana Labs.

Broader Market Signals

Oxide’s raise is the largest single‑handed AI‑hardware bet since Nvidia’s 2022 $1.5 billion infusion into its Arm‑based platform. It confirms that venture capital is now willing to fund full stack hardware, not just software AI layers. The involvement of Andreessen Horowitz, a firm that recently backed Cerebras and Graphcore, suggests a belief that custom silicon will become a commodity for data‑center operators. The round also puts pressure on incumbents to open their designs; Oxide’s open‑source OS, Redox, could force cloud giants to adopt more transparent firmware to stay competitive.

Oxide now stands at a crossroads: deliver a production‑grade silicon design on schedule, or watch its runway evaporate as rivals outpace it. The $445 million war chest gives the startup a realistic shot at disrupting a market dominated by megacorp engineers. Success would validate a new model of open‑hardware, venture‑backed cloud infrastructure. Failure would reinforce the entrenched belief that only the biggest players can afford the cost of custom silicon. The industry will be watching the next silicon tape‑out like a stock ticker.

Sources: Oxide Computer blog post (https://oxide.computer/blog/our-445m-series-d), Crunchbase funding data, press releases from Andreessen Horowitz and Sequoia Capital.