Skip to content

Software Engineer - Software Engineer - GPU, C++, OpenCL, CUDA

  • On-site
    • Waukesha, Wisconsin, United States
  • ENGINEERING

Job description

Position: Software Engineer - GPU, C++, OpenCL, CUDA

Location: Waukesha, WI (Onsite)

Exp: 5 – 9 yrs

Key Skills: GPU, C++, OpenCL, CUDA, OneAPI, Matlab

 

Job Requirements
The CT Program is working on upgrading CT scanners used worldwide. The center is currently concentrating on the ongoing enhancement of the next generation of CT machines, including their essential workflows and applications. For that purpose, proficient and experienced resources are required.

Primary Objective:

  1. Leverage proprietary software platform to implement image processing algorithms on GPUs. (C++/OpenCL/CUDA/OneAPI)

  2. Improve image chain performance using heterogeneous high-performance computing (HPC) to meet customer expectations

  3. Ensure quality and compliance of productized code per regulatory expectations

Detailed Requirements:

  1. Productized CT image processing algorithms on GPU, including ported algorithms from Matlab to GPU, or OpenCL to CUDA

  2. Improved image chain & algorithm performance compared to initial benchmarks

  3. Perform GPU profiling, identify algorithm bottlenecks, troubleshoot and resolve performance issues

  4. Improve GPU utilization leveraging heterogenous HPC knowledge.

  5. Perform testing, reliability analysis, performance benchmarks and document results

  6. Execute test procedures with high quality and rigor, following Good Documentation Practices


Work Experience
Skills:

  1. Programming Languages: C++, OpenCL, CUDA, OneAPI

  2. Image Processing Algorithms: Implementation and optimization on GPUs

  3. High-Performance Computing (HPC): Knowledge of heterogeneous HPC

  4. Profiling and Performance Analysis: GPU profiling, identifying bottlenecks, troubleshooting, and resolving performance issues

  5. Testing and Documentation: Performing testing, reliability analysis, performance benchmarks, and documenting results following Good Documentation Practices

Additional Experience:

Productizing Algorithms: Experience in productizing CT image processing algorithms on GPU
Porting Algorithms: Experience in porting algorithms from Matlab to GPU or OpenCL to CUDA
Improving Performance: Proven track record of improving image chain and algorithm performance compared to initial benchmarks
Quality and Compliance: Ensuring quality and compliance of productized code per regulatory expectations

Best regards,


Prasad Kalsekar | Hudson Manpower
Email: prasad@hudsonmanpower.com

or