Device Driver Engineer
Core
Design and implement high-performance driver stacks for AI accelerator hardware to enable low-latency, high-bandwidth communication between host systems and edge-to-cloud AI chips.
Role type
Device Driver Engineer (Linux/Windows)
Builds
Production device drivers and kernel modules for AI accelerators
Domain
AI hardware / Edge computing / Semiconductor systems
Deliverable
production ML models
Required skills
Linux/Windows device driver development, PCIe protocol implementation, C/C++ kernel-mode programming, DMA and interrupt handling, memory management, low-level debugging (gdb, perf, ftrace), multi-threading and synchronization
Preferred skills
AI accelerator experience (GPUs, NPUs, TPUs), virtualization (VFIO, SR-IOV, DPUs), containerized environments (Kubernetes, Docker), secure boot and TEE
Technologies
Linux Kernel, Windows WDDM/WDF/MCDM, PCIe Gen4/Gen5, SR-IOV, BAR, IOMMU, VFIO, Kubernetes, Docker, gdb, perf, ftrace, dmesg
Responsibilities
Develop and optimize Linux/Windows PCIe device drivers for AI accelerators; Implement low-level hardware interactions including DMA, memory management, and interrupt handling; Debug and troubleshoot PCIe protocol issues, kernel panics, and performance bottlenecks; Collaborate with hardware, firmware, and AI software teams to define driver interfaces; Ensure compliance with PCIe standards and support virtualization/containerized environments; Develop tools for profiling and monitoring driver performance
Seniority
Mid-level, hands-on IC
