We’re looking for an experienced datacenter virtualization engineer with GPU or accelerators expertise, possibly datacenter Engineer to help with the development of a cloud-based, ARM and x86-compatible virtual environment with GPU sharing capability. This role will focus on configuring, optimizing, and managing an emulated infrastructure to support our client's hardware, enabling seamless development and testing in a fully virtualized environment.
Key Responsibilities:
- Help in the configuration, optimization, and management of bare metal AWS cloud instances to support emulation for both ARM and x86 with GPU capabilities.
- Develop multi-architecture, cross-compiled containers, utilizing Docker Buildx and GCC toolchains, Qemu, or other solutions to support GPGPU for emulation and facilitate scalability.
- Collaborate closely with software and hardware teams to ensure all drivers, applications, and dependencies are optimized for a hybrid ARM/x86 emulated environment.
- Establish and manage container orchestration (e.g., K3s, Kubernetes) for OCI-compliant containers, focusing on efficient GPU access and seamless cross-platform compatibility.
- Provide guidance on ARM and x86 cross-compilation practices, ensuring performance, security, and scalability across the virtualized development lifecycle.
- Work with network engineers to integrate SDWAN within the emulation environment, meeting high-bandwidth, low-latency requirements.
- Enhance security and compliance by hardening cross-architecture container environments, specifically around GPU access and data handling.
Qualifications:
- 5+ years of experience with virtualization architecture, multi-architecture containerization, and cross-compilation.
- Strong expertise with AWS services, particularly those supporting GPU (e.g., AWS Graviton, G4DN, bare metal instances is needed).
- Proficiency with cross-compilation toolchains (GCC, LLVM) and in building OCI-compliant multi-architecture containers.
- Experience with NVIDIA GPU Enterprise drivers, CUDA libraries, and managing GPU access in containerized setups.
- Skilled in K3s, Kubernetes, or similar container orchestration solutions.
- Familiarity with SDWAN configurations and best practices for hybrid cloud networks building emulators (e.g. ARM on X86) or multi instance networking.
- Advanced understanding of Linux-based systems, including low-level hardware interactions Linux kernel compilation, and device drivers.
Preferred Qualifications:
- Background in high-performance computing (HPC) and managing GPU-accelerated workloads.
- Knowledge of security best practices in virtualized and containerized environments.
- Previous experience with multi-cloud or hybrid cloud deployments is a big plus.
- Can turn ambiguity to solution