Overview
We are seeking a highly skilled Linux Kernel/Driver Developer with strong expertise in C/C++ programming to join as a KFD Driver Development and Debugging Engineer. The role requires a balanced skill set (50:50 development and validation), with a strong focus on Linux kernel development, GPU driver debugging, and performance optimization. You will be responsible for developing, debugging, and optimizing the Kernel Fusion Driver (KFD) for GPUs.
- Develop and maintain the KFD driver, ensuring seamless integration with the ROCm stack.
- Debug, reproduce, and resolve driver bugs, crashes, and performance issues.
- Collaborate with hardware, firmware, and software teams to ensure GPU driver stability and integration.
- Use tools such as kgdb, ftrace, and tracepoints for kernel-level debugging and root cause analysis.
- Conduct performance profiling, tuning, and optimization of GPU drivers.
- Write and maintain technical documentation for driver functionality, usage, and troubleshooting.
- Participate in code reviews and contribute to best practices in Linux kernel programming.
- Stay updated with the latest advancements in Linux kernel and GPU technologies.
Required Skills & Qualifications
- B.E/B.Tech or M.E/M.Tech in Computer Science, Computer Engineering, Electronics, or related field.
- 3+ years of experience in Linux kernel development or driver programming (preferably GPU drivers).
- Strong understanding of Linux kernel architecture, device drivers, and low-level programming concepts.
- Proficiency in C programming, debugging methodologies, and performance tuning on Linux.
- Familiarity with GPU architecture and the ROCm ecosystem.
- Hands-on experience with memory management, synchronization, and concurrency in multi-threaded environments.
- Strong analytical and problem-solving skills.
- Excellent communication and documentation abilities.
Preferred Skills
- Experience with Kernel Fusion Driver (KFD) or GPU-related driver development.
- Prior contributions to open-source kernel/driver projects.
- Familiarity with performance tuning and profiling tools for GPU workloads.