AI System Design Engineer (Internship)

OMNIVISION; Singapore; On-site; SGD 27,900 - 39,060; Full time

Job summary

OMNIVISION in Singapore is seeking a system design engineer intern focused on AI compute. You will join an R&D team to shape the AI compute roadmap for sensor and deeply embedded processor systems.

The role emphasizes hands-on work in system modelling, prototyping, and validation, inviting you to explore the complexities of performance and power efficiencies. Candidates should possess an MSEE/BSEE and experience with Python and C/C++, along with a drive for excellence.

Qualifications

• Disciplined, quality-minded, and highly driven for excellence.
• Familiarity with ARM or RISC-V is a plus.
• Exposure to deep learning (e.g., CNNs, Transformers) is a plus.

Responsibilities

• Work on system modelling and profiling.
• Conduct rapid prototyping and system validation.
• Understand trade-offs in performance, power, and area.

Skills

Team player

Communication skills

Python

C/C++

Basic understanding of computer architecture

Education

MSEE/BSEE in Electrical Engineering or Computer Engineering

Tools

Verilog

SystemVerilog

Embedded systems knowledge

Job description

As a system design engineer intern for AI compute, you will join a small R&D team focused on shaping the AI compute roadmap for sensor and deeply embedded processor systems. The team works across SW/HW boundaries to explore and prototype efficient AI architectures under real-world constraints.

You will work on one or more areas including system modelling and profiling, roofline and bottleneck analysis, rapid prototyping, and system validation. The work is hands‑on and exploratory, with an emphasis on understanding trade‑offs in performance, power, and area.

Requirements

• Disciplined, quality‑minded, and highly driven for excellence
• Excellent team player and good communication skills
• MSEE/BSEE in Electrical Engineering or Computer Engineering
• Basic understanding of computer architecture and memory hierarchies
• Familiar with Python and C/C++
• Exposure to deep learning (e.g., CNNs, Transformers) is a plus
• Exposure to NPU / GPU / DSP is a plus
• Familiarity with ARM or RISC‑V is a plus
• Some knowledge of Verilog or SystemVerilog is a plus
• Exposure to embedded systems is a plus
• Awareness of low‑precision formats (e.g., MX, Q4_K, INT8, FP16) is a plus
• Commit at least 6 months full time and preferably 9 months

Hands‑on experience with chip modelling, performance simulation, or AI compute systems will be an advantage. You should be comfortable working with incomplete problems and helping define the solution space, not just implementing a fixed spec.