GPU Architecture and Performance Expert

Description:

About the Company

A global leader in electronic design automation software, this organization empowers innovators in the microchip and semiconductor industries to push the boundaries of technology. With cutting-edge solutions that enable optimal circuit design and semiconductor development, the company serves as a trusted partner to organizations driving the future of electronics and computational hardware.

About the role

We are seeking a GPU Architecture and Performance Expert to assess and advise on GPU porting strategy for our product, specifically evaluating whether GPU acceleration can deliver material performance improvements for our integer-based algorithms.

Responsibilities

Conduct comprehensive technical assessment of the product's native algorithms to determine viability and efficiency gains from GPU acceleration

Analyze integer-based computational workloads and specialized solvers to evaluate whether GPU architectures can provide meaningful performance improvements

Provide strategic recommendations on GPU investment decisions, including cost-benefit analysis and expected acceleration potential for specific algorithm types

Deliver data-driven insights on which computations are strong candidates for GPU porting versus those that should remain CPU-optimized

Guide technical teams on implementation strategies if GPU acceleration proves viable, including architecture selection and optimization approaches

Required Skills (All Must Haves)

Deep expertise in GPU architectures (CUDA, OpenCL, or similar) with proven experience evaluating workload suitability for GPU acceleration

Strong understanding of the performance characteristics of integer-based algorithms on GPU versus CPU architectures

Demonstrated experience analyzing and optimizing computational workloads for heterogeneous computing environments

Proficiency in performance profiling, benchmarking, and cost-benefit analysis for hardware acceleration decisions

Solid foundation in parallel computing principles and understanding of when GPU acceleration provides material benefits

Open Positions: 1

Detailed Required Skills: Algorithms, Architecture, GPU Computing, NVIDIA CUDA

Job Timezone: America/New_York

Work Type: Remote

• Estimated Length: 4-8 weeks