A

Arkane Cloud

Cloud GPU servers for AI training

Vice President Infrastructure

Platform EngineeringUpdated on Nov 13
Full-timeLyon, France

About the Job

Requirements

  • At least 10 years of infrastructure experience, including at least 3 years leading a GPU infrastructure team.
  • Personally delivered at least one NVIDIA HGX cluster with several hundred GPUs.
  • Expertise in high-performance networking using InfiniBand or RoCE v2, including rail-optimized and fat-tree designs and fabric management.
  • Data center expertise, including colocation, electrical and thermal constraints, and liquid cooling or more than 120 kW per rack.
  • Experience negotiating hardware and colocation contracts.
  • Production operations experience covering service-level agreements, on-call duties, and incident management.
  • Professional fluency in English.

Responsibilities

  • Design and deploy GPU clusters according to NVIDIA reference architectures.
  • Own networking and storage, including InfiniBand XDR, RoCE v2, external connectivity, and VAST Data.
  • Select and negotiate with data centers and suppliers, including colocation, energy, liquid cooling, NVIDIA, original equipment manufacturers, and integrators.
  • Maintain service-level agreements through monitoring, on-call operations, incident response, and automation, and deliver the infrastructure component of ISO 27001 and then SecNumCloud.
  • Build and manage the infrastructure and DevOps team, and oversee total cost of ownership and revenue per GPU.

Desired Qualifications

  • Experience at a cloud provider, GPU-focused cloud provider, or high-performance computing company.
  • Kubernetes and GPU orchestration experience, including GPU Operator, MIG, and Slurm.
  • Infrastructure as code experience.
  • High-performance storage experience.
  • Inference stack experience, including vLLM and SGLang.

About the Company

Arkane Cloud

Cloud GPU servers for AI training

Arkane Cloud builds and manages GPU-backed cloud infrastructure for Artificial Intelligence. It provides GPU servers and related cloud resources so customers can train AI models and run high-scale inference at scale. The service centers on offering dedicated capacity for AI workloads, enabling users to provision powerful GPUs as needed without managing the hardware themselves. This specialization in AI-focused GPU deployment differentiates Arkane Cloud from more general cloud providers, emphasizing scalable resources tailored for training and large-volume inference with the goal of helping customers accelerate AI development and deployment.

Company size
Not listed
Company stage
Not listed
Total funding
Not listed
Headquarters
Lyon, France
Founded
Not listed

Recently Posted Jobs

Explore more opportunities at Arkane Cloud

See more jobs