Forward-Deployed AI Engineer: GenAI Inference on AMD Hardware

AMD; Medina; In loco; SAR 260.000 - 460.000; Tempo pieno

Descrizione del lavoro

AMD is seeking a highly skilled Member of Technical Staff to join the Forward Deployed Engineering team. You will work directly with strategic partners to optimize and deploy GenAI inference workloads on AMD Instinct GPUs, bridging AI models and hardware for maximum performance.

In this customer-facing role, you will optimize pipelines, develop low-level kernels, and contribute to open-source frameworks while leveraging AI-assisted tooling to accelerate delivery and ensure accuracy and

Competenze

• BS/MS/PhD in Computer Science or Electrical Engineering or a related field.
• 8+ years of relevant experience.
• Strong GPU architecture and performance optimization experience (ROCm or CUDA).
• Proficiency in Python and C/C++, kernel development using Triton/HIP/CUDA.
• Experience with LLM inference frameworks (vLLM, SGLang or equivalent) and serving optimization techniques.
• Understanding of transformer and MoE architectures, quantization and distributed inference strategies.

Mansioni

• Partner with customers as part of the FDE team to enable and optimize large language models and multimodal models on AMD Instinct GPUs, meeting performance and accuracy targets.
• Optimize GenAI inference pipelines with vLLM and SGLang including disaggregation, data-parallelism, and KV cache management.
• Develop and tune low-level GPU kernels using Triton, HIP, CUDA.
• Profile end-to-end workloads and identify bottlenecks across GPU-CPU boundaries for throughput/latency improvements.
• Enable advanced data types and techniques (FP8, MXFP4, sparse attention) with emphasis on performance and numerical accuracy.
• Debug runtime stability and accuracy issues in production and contribute fixes with core engineering teams.
• Contribute optimizations to open-source frameworks (vLLM, SGLang, ATOM, AITER) and document best practices.

Conoscenze

GPU architecture

Python

C/C++

LLM inference frameworks

Transformer/MoE architectures

Parallelism TP/DP/EP

Profiling tools

AI-assisted coding tools

Formazione

BS in Computer Science or Electrical Engineering

MS/PhD in Computer Science or Electrical Engineering

Strumenti

ROCm

CUDA

Triton

HIP

GitHub Copilot

Claude

Cursor

Descrizione del lavoro

AMD is seeking a highly skilled Member of Technical Staff to join the Forward Deployed Engineering team. You will work directly with strategic partners to optimize and deploy GenAI inference workloads on AMD Instinct GPUs, bridging AI models and hardware for maximum performance.

In this customer-facing role, you will optimize pipelines, develop low-level kernels, and contribute to open-source frameworks while leveraging AI-assisted tooling to accelerate delivery and ensure accuracy and