FuriosaAI is looking for outstanding software engineers to design and build an end-to-end Kubernetes stack for next-generation AI chips.
In this role, you will work closely with the various teams for kubernetes integration, design, and build a resource scheduling and management solution at datacenter scale as part of a small team of experienced software engineers.
We are focused on supporting our products as first-class resource in kubernetes accross a variety of customer environments, both bare-metal and virtualized platforms. We expect you to have strong programming skills related to Go or Rust and deep understanding of kubernetes and linux container internals as well as excellent communication and planning skills.
Integrate FuriosaAI NPU products as first-class resources in Kubernetes ecosystem.
Develop and maintain scheduling and resource management solution for datacenter scale.
Collaborate with teams working on ML applications, SDK, runtime, device driver, firmware and infrastructure.
Maintain high-quality software products through robust test coverage.
Bachelor’s degree in Computer Science/Engineering, or any related field, or a equivalent work experience
Strong understanding of Data Structures and Algorithms.
Proven ability to write high-quality, efficient, and maintainable code.
3+ years of strong experience with the Go or Rust programming language.
Solid understanding of Kubernetes Architecture, Container Runtime and Linux container internals.
Background in Cloud Native environment and distributed system design.
Experience of writing dynamic GPU Scheduler for MIG-Capable GPUs.
Knowledge of Hardware Topology Management and Optimization for NUMA or PCIE-Topology or NVLink/NVSwitch
Real world strong hands-on experience of Building and Operating GPU Farm or GPU Cluster.
Experience of writing Kubernetes Operator with kubebuilder, operator sdk or informer.
