NPU Library Software Engineer (AI)

TL;DR

NPU Library Software Engineer (AI): Designing and developing compute libraries for neural network operations optimized for proprietary NPU instruction set architecture. Focus on improving utilization of heterogeneous compute resources and enhancing functional coverage through precision management and numerical stability.

Location: Onsite in Seongnam, South Korea

Company

hirify.global is an AI semiconductor company developing specialized hardware for high-performance machine learning workloads.

What you will do

• Design and develop compute libraries such as BLAS and DNN libraries.
• Optimize neural network operations for proprietary NPU architecture.
• Manage tensor shape variations, precision, and numerical stability.
• Improve performance utilization of heterogeneous compute resources based on workload characteristics.

Requirements

• Master’s degree or higher in Electrical Engineering, Computer Science, or related field.
• In-depth understanding of neural network operations and computational workflows.
• Strong analytical, troubleshooting, and performance optimization skills.
• Proficiency in C++ and Python.

Nice to have

• Broad knowledge of deep learning models including computer vision, NLP, and speech recognition.
• Experience with model and layer-level optimization techniques.
• Experience with hardware acceleration frameworks like CUDA, OpenCL, SSE, AVX, or NEON.
• Academic background in computer architecture.

Hiring process

• Application review followed by online and onsite interviews.
• Onsite interview includes a technical task.
• Final culture-fit interview and offer negotiation.