Graduate Talent (vdi & Media AI Systems Engineering)

Intel Intel · Semiconductors · Penang, Malaysia

This role focuses on optimizing and evaluating AI workloads, particularly for VDI and Media AI systems, on Intel Xeon CPU platforms and AI accelerators. Key responsibilities include benchmarking inference performance, supporting AI PoCs, enabling GPU-accelerated inferencing, and creating automation pipelines for AI service deployment. The role involves working with AI frameworks and AI workloads like Vision-Language Models, Generative AI, and Agentic AI, with a focus on inference performance and resource utilization.

What you'd actually do

  1. Optimize VDI performance on server CPU platforms via automated benchmarking and performance analysis.
  2. Develop and evaluate AI workloads in cloud AI PC environments on inference performance, and CPU/GPU resource utilization.
  3. Support development of Media AI proofs-of-concept (PoC), enable server CPU and workstation/data center GPU-accelerated AI inferencing, advanced content creation, rendering, media processing and video analytics.
  4. Create automation, data collection, and AI service pipelines while debugging system and application issues across distributed environments.
  5. Professional documentation to communicate performance insights, support engineering decisions and reproducible testing.

Skills

Required

  • Degree in Computer Science, Computer/Electrical Engineering, or related field.
  • Strong programming skills in Python or C/C++.
  • Basic knowledge of Linux and Windows, virtualization (VDI, hypervisors), system performance components and optimization concepts (CPU, memory, I/O).

Nice to have

  • Experience with software development and benchmarking
  • AI/ML frameworks (PyTorch, TensorFlow, Transformers)
  • AI workloads (Vision-Language Models, Generative AI, Agentic AI)
  • GPU programming or CUDA-based workloads
  • Familiarity with cloud platforms (AWS, Azure)
  • Containerization and APIs (REST, MCP, service-based architectures)
  • Knowledge of Intel architecture (Intel Xeon, Intel AMX) and profiling tools (e.g., Intel VTune Profiler)

What the JD emphasized

  • AI accelerators
  • GPU-enabled workloads
  • AI inferencing
  • Generative AI
  • Agentic AI

Other signals

  • AI accelerators
  • GPU-enabled workloads
  • AI PC-in-the-cloud solutions
  • AI workloads
  • AI inferencing
  • Generative AI
  • Agentic AI