Senior Software Engineer - Compute Performance
Fremont , CA 94537
Posted 2 months ago
This Job is not relevant Tell us why
At Pony.ai, excellence is at the core of our company. Every detail counts as we work to deliver a safe, reliable, and cost-effective autonomous driving platform to the public. As part of the team, you will play a critical role in developing and deploying state-of-the art vehicle technology. You will have every opportunity to grow and develop as our company grows. Join us on this exciting journey as we bring the latest breakthroughs in artificial intelligence to the future of transportation.
As a Senior Software Engineer in compute performance, you will be developing technologies to accelerate leading applications in autonomous driving fields within computer vision/machine and deep learning/mapping and localization/planning and prediction, performing in-depth analysis and optimization to ensure the best possible performance on current CPU/GPU/FPGA architectures.
- BS/MS or Ph.D in computer science or electrical engineering a related field.
- Strong programming skills in C/C++.
- Good knowledge of the fundamentals of computer architecture, including CPU data caching, SIMD instruction sets, and compiler code generation.
- Good understanding of hardware performance is required, preferably regarding CPU and GPU execution model, threads, registers, cache, cost/performance trade-off, etc.
- Experience with benchmarking and validating performance for complex computing architectures.
- Experience in optimization utilization of compute resources, identifying and resolving data flow bottlenecks.
- Experience with parallel programming, ideally CUDA or OpenCL.
- Strong communication skills and ability to work cross-functionally between software and hardware teams
- Strong knowledge of software design, programming techniques, and algorithms is prefered.
- Strong mathematical fundamentals, including linear algebra and numerical methods is prefered.
- Experience in computer vision, machine learning and deep learning is preferred.