# Google introduces Gemini 4 Argon in a limited cybersecurity rollout

Google introduced Gemini 4 Argon to a small group of cybersecurity partners. HPE said Vultr placed a $1.2 billion order for HPE systems built around AMD's Helios AI rack architecture, to be deployed at Vultr data centers in the United States.

Topics: models, safety, infrastructure

Coverage: 2026-09-29T21:00:00.000Z to 2026-09-30T21:00:00.000Z

## Google introduces Gemini 4 Argon in a limited cybersecurity rollout

Google introduced Gemini 4 Argon to a small group of cybersecurity partners. The company describes it as a model for long-running work in areas including software engineering and cybersecurity, and says it leads on selected coding and knowledge-work benchmarks. Google plans to broaden access after further testing. 

This is Google's next-generation flagship announcement, but the staged release means the immediate development is access for a select group—not a broadly available model. The rollout makes Google's claims and future independent testing worth watching.

Gemini 4 Argon is initially limited to selected cybersecurity partners, not generally available. Benchmark comparisons are Google's claims; independent evaluations are not yet available, and Axios reported conflicting accounts of internal test performance.

[Google unveils long-awaited Gemini 4](https://www.axios.com/2026/09/30/google-gemini-4)

## Vultr places $1.2 billion order for HPE's AMD Helios AI racks

HPE said Vultr placed a $1.2 billion order for HPE systems built around AMD's Helios AI rack architecture, to be deployed at Vultr data centers in the United States. HPE says this is its first Helios order; each rack integrates 72 AMD Instinct MI455X GPUs alongside CPUs, networking, and ROCm software. 

The deal is a substantial customer commitment to AMD-based rack-scale AI infrastructure and gives HPE a first Helios deployment order. It adds a concrete infrastructure development to a news cycle otherwise dominated by model and policy announcements.

HPE announced the order on September 30, before coverageStart; the source does not give a precise announcement time. The deployment is a future plan, and the $1.2 billion figure is a company-reported order, not evidence that the capacity is already operating.

[HPE secures its first AMD Helios order in $1.2 billion deal with Vultr](https://investors.hpe.com/?ReleaseID=909541)
