Senior Dl Algorithms Engineer - Inference Performance

NVIDIA NVIDIA · Semiconductors · Santa Clara, CA +1 · Remote

Senior DL Algorithms Engineer focused on optimizing inference performance for language and multimodal models using NVIDIA's inference stack (NIMs, TRT-LLM). Role involves profiling, analysis, and collaboration across hardware/software layers to maximize performance on GPUs.

What you'd actually do

  1. Implement language and multimodal model inference as part of NVIDIA Inference Microservices (NIMs).
  2. Contribute new features, fix bugs and deliver production code to TRT-LLM, NVIDIA’s open-source inference serving library.
  3. Profile and analyze bottlenecks across the full inference stack to push the boundaries of inference performance.
  4. Benchmark state-of-the-art offerings in various DL models inference and perform competitive analysis for NVIDIA SW/HW stack.
  5. Collaborate heavily with other SW/HW co-design teams to enable the creation of the next generation of AI-powered services.

Skills

Required

  • PhD in CS, EE or CSEE or equivalent experience
  • 5+ years of experience
  • Strong background in deep learning and neural networks, in particular inference
  • Experience with performance profiling, analysis and optimization, especially for GPU-based applications
  • Proficient in C++, PyTorch or equivalent frameworks
  • Deep understanding of computer architecture, and familiarity with the fundamentals of GPU architecture

Nice to have

  • Proven experience with processor and system-level performance optimization
  • Deep understanding of modern LLM architectures
  • Strong fundamentals in algorithms
  • GPU programming experience (CUDA or OpenCL)

What the JD emphasized

  • performance analysis and optimization
  • squeeze every last clock cycle
  • across all layers of the hardware/software stack
  • peak performance
  • inference performance
  • GPU-based applications
  • processor and system-level performance optimization

Other signals

  • NVIDIA Inference Microservices (NIMs)
  • TRT-LLM
  • inference performance
  • GPU architecture
  • deep learning frameworks