Celestica is looking for talented and passionate software engineers to help drive the next generation data center and AI compute blades, and to be responsible for designing, developing and testing the software that drives high density data center compute racks including GPUs, CPUs and AI accelerators.
Requirements
- Proficiency in C, C++ and other low-level programming languages.
- Experience with Linux driver kernel development is a must. Candidate should have strong experience in Linux kernel memory management, interrupts, drivers etc
- Familiarity with main elements of CPU, DPU, memory, NICs, board monitoring elements is a must
- Ability to identify and resolve software and hardware issues at the rack level.
- Strong analytical and problem-solving abilities
- Prior experience in data center architectures, developing and maintaining software for deployments is a plus
- Prior experience GPU libraries like CUDA and ROCm is a plus
Responsibilities
- Design and implement drivers, application software that drive next generation CPU and GPU
- Develop, augment and maintain software libraries for computing on GPU
- Optimize code for performance, efficiency, and low latency
- Innovate new ideas in conjunction with networking engineers to optimize overall data center efficiency
- Board bringup and validation of all low-level interactions with hardware
- Drivers to monitor the health of the board
- Troubleshoot and resolve hardware-software interaction issues
Other
- Work effectively in a team environment, collaborating with engineers and peer functional leads from different disciplines to innovate solutions, triage issues and speed execution
- Explain the value addition the software brings to technical and non-technical audiences.
- Lead the junior teams to solve technical problems.
- Review and maintain quality of code going into production.
- Architect software applications that will work across multiple technology generations with minimal changes.
