CUDA Developer (100% Remote)

Customer - A venture backed company is implementing a pivot. The firm wants to get an unbiased report on the existing codebase. Our output would be to produce documentation on observations on codebase, thru codereviews, for performance, optimization and general alignment with the Data Center Observability area.

Job Description: CUDA Developer (100% Remote)

Location: Remote

Job Type: Part-time consulting - 10-15 hours a week

About the Role: We are seeking a highly skilled and motivated Software Engineer (3-5 years of experience) to join me working on a Data Center Observability and Analytics Platform. In this role, you will be responsible for code reviews, optimizations, and contributions to existing CUDA-based codebase to enhance the platform's performance, scalability, and efficiency. This is not a development job but only code review, writing a report for code flow, report on optimal and efficiencies of the codebase [there is no code to write].

Responsibilities:

• Code Reviews: Conduct thorough code reviews for CUDA-based implementations to ensure high performance, reliability, and adherence to best practices.

• CUDA Optimization: Suggest optimization and efficiency steps for the CUDA code to accelerate analytics, observability, and data processing within the platform.

• Performance Tuning: Focus on optimizing data processing pipelines using GPU acceleration, ensuring low latency, high throughput, and scalability across large-scale data centers.

• Platform Improvements: Identify and implement performance bottlenecks in the existing codebase and suggest architectural improvements related to GPU acceleration.

• Documentation: Provide clear documentation for the CUDA codebase and any optimizations or architectural changes made to the platform.

Requirements:

• Experience: 3-5 years of experience as a Software Engineer with hands-on experience in CUDA programming and GPU-based optimizations.

Technical Skills:

• Strong proficiency in C/C++ programming, with a focus on performance-critical applications.
• Solid experience with CUDA development and optimizing code for GPU performance.
• Familiarity with GPU libraries like cuBLAS, cuDNN, or similar, as well as tools for profiling and debugging CUDA code.
• Experience with data center observability, monitoring, or analytics platforms is a plus.
• Proficient with parallel computing concepts and multi-threading in GPU and CPU environments.