Kernel Driver Software Engineer

About the role

About Etched

Etched is building hardware for frontier intelligence. We co-design chips, racks, software, and manufacturing to deliver best-in-class throughput and latency across both prefill and decode workloads. Our first products are heavily focused on inference. Backed by hundreds of millions from top-tier investors and staffed by leading engineers, Etched is redefining the infrastructure layer for the fastest growing industry in history.

Key Responsibilities

- Design, develop, and maintain kernel-mode drivers ensuring high reliability, informative debug, and optimal performance.

- Analyze and optimize driver performance for demanding AI workloads, focusing on minimizing latency and maximizing throughput.

- Collaborate closely with hardware engineers throughout the ASIC design process..

- Implement driver support for device virtualization technologies, including SR-IOV, VFIO, and para-virtualization.

- Implement efficient memory management strategies considering kernel memory mapping, page tables configuration, NUMA awareness for device data caching, and IOMMU configuration.

- Build kernel drivers fundamentally designed to support and maintain security across host processes, physical memory spaces, and device attestation.

- Diagnose and resolve complex driver-related issues, using common kernel debugging tools and techniques (ftrace, dmesg, etc.) to identify and fix bugs.

- Design and implement synchronization mechanisms to handle concurrent access to  multiple accelerators.

- Develop and execute comprehensive test plans to validate driver functionality, stability, and performance in manufacturing and in general production environments.

- Collaborate with software and hardware teams to diagnose and resolve complex system-level issues.

Representative Projects

- Develop and optimize kernel-mode drivers for new ML accelerators.

- Implement and optimize memory management, including kernel memory mapping and IOMMU configurations, for high-bandwidth data transfers.

- Debug and resolve complex driver-related issues impacting ML workload performance.

- Develop performance benchmarks and profiling tools to analyze driver performance.

- Integrate driver support for advanced features like hardware virtualization and security, including SR-IOV and VFIO.

- Optimizing PCIe communication between the host and PCIe devices, using advanced equipment like PCIe analyzers.

- Implement and debug power management features for PCIe devices.

- Integrating ML accelerators into containerized and virtualized environments.

- Implementing and optimizing para-virtualization techniques for PCIe devices.

- Configure and optimize page tables for efficient memory access from the ML accelerator.

- Participate in hardware-software co-design reviews across teams to optimize performance and power efficiency.