(204) Machine Learning Software Engineer (GPU / ROCm / CUDA / C++ / Python)

No. of Positions : 1

Experience : 3+ Years

Location : Bangalore

Key Skills / Highlights : C/C++, Python, ML frameworks, ROCm/CUDA, Linux, GPU architecture

Urgency : High

JD

• MS/BS degree in Computer Science or an equivalent.
• Experience with Linux Commands is must.
• Experience with Scripting language like bash/powershell.
• Understanding of various python ML frameworks like Pytorch, Transformers etc.
• Understanding of various language and compiler for writing highly efficient custom Deep-Learning GPU Kernels. like Triton/Jax.
• Hands on Debugging Experience with gdb, valgrind etc.,br> • Experience and understanding of AI Models and Inferencing Engines like vllm/ollama/llama.cpp/sglang.
• Experience with Profiling tools needed to debug CUDA/ROCm Kernels like nsys/rocprof is a plus.
• Knowledge of GPU architecture, PC architecture.
• Experience in writing ROCM/CUDA Kernels/Shader.
• Deep understanding and experience in implementation of Machine learning and AI algorithm.
• Good communication skills and able to work with stakeholders effectively.
• Knowledge of x86 assembly language and x86/x64 CPU instructions is a plus.

Responsibilities

• Work on latest machine learning technologies.
• Work on supporting for latest Linux operating system.
• Work on AMD next generation GPUs/Accelerators.
• Work on optimizing latest Rocm drivers and improve performance.
• Design new machine learning technologies.