GPU Virtualization Engineer @ Innowise

Must-Have Requirements:

• Strong systems programming proficiency in C and/or Rust.
  
• Deep, hands-on experience with KVM/QEMU, including patching, extending, or debugging the hypervisor layer itself (not just operating it).
  
• Solid understanding of Linux kernel internals (specifically memory management, device drivers, or the VFIO subsystem).
  
• Practical experience with at least one of the following: live migration internals, CRIU (Checkpoint/Restore In Userspace), NVIDIA vGPU/VFIO passthrough, or libvirt device modeling.
  
• Comfort working in and collaborating with upstream open-source communities (such as QEMU, KubeVirt, libvirt, or CRIU).
  
• English B1+

Nice to Have:

• Proven upstream commits in QEMU, KubeVirt, libvirt, CRIU, or Kata Containers.
  
• Experience with advanced networking and memory optimizations, including GPUDirect RDMA, NCCL, RoCE/InfiniBand, or complex PCIe-topology configurations.
  
• Familiarity with KubeVirt/Kata-based VM runtimes inside Kubernetes and handling CUDA state serialization.
  
• Prior experience working on core platforms at industry leaders (e.g., NVIDIA, Red Hat, VMware, hyperscaler VM-infrastructure teams, or GPU neo-clouds like CoreWeave/Lambda).

Requirements: C, Rust, QEMU, KubeVirt