Staff Cloud Hypervisor Architect for AI Infra

Crusoe; Sunnyvale (CA); Presencial; USD 215.000 - 260.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Competitive compensation

Equity packages

Generous PTO

Health insurance

401(k) match

Professional development

Descripción de la vacante

Crusoe is building the engine room of the AI revolution and seeks a Staff Cloud Hypervisor R&D Engineer to architect a next-generation virtualization stack. You will lead core hypervisor components optimized for massive-scale GPU fleets and enable near-zero-latency hardware pass-through for GPUs and DPUs in a multi-tenant cloud.

You will tackle live migration, PCIe interconnect optimization, and high-performance I/O paths while contributing to upstream virtualization projects.

Formación

• 7+ Years of Experience in hypervisor internals, kernel development, or low-level systems programming.
• Deep virtualization expertise including CPU virtualization and memory virtualization.

Responsabilidades

• Lead the R&D and implementation of core hypervisor components (KVM, QEMU, or Rust-based) optimized for large GPU fleets.
• Develop hardware pass-through and abstraction techniques (SR-IOV, VFIO, mdev) for multi-tenant NVIDIA GPUs and DPUs.
• Solve live migration and PCIe peer-to-peer challenges for AI workloads.
• Conduct bottleneck analysis across CPU, MMU, and guest OS scheduling to maximize throughput.
• Contribute to upstream open-source virtualization projects and drive security boundaries.

Conocimientos

Hypervisor internals

Virtualization

C/C++

Rust

QEMU/KVM

Linux kernel debugging

PCIe/I-O paths

Security/isolation

Herramientas

QEMU

KVM

Linux kernel

Perf

eBPF

Descripción del empleo

Crusoe is building the engine room of the AI revolution and seeks a Staff Cloud Hypervisor R&D Engineer to architect a next-generation virtualization stack. You will lead core hypervisor components optimized for massive-scale GPU fleets and enable near-zero-latency hardware pass-through for GPUs and DPUs in a multi-tenant cloud.

You will tackle live migration, PCIe interconnect optimization, and high-performance I/O paths while contributing to upstream virtualization projects.