17 hours ago
Bengaluru, IndiaSenior / Staff+
Responsibilities
- Benchmark and characterize AI inference workloads across CPU, GPU, APU, NPU, and heterogeneous computing platforms.
- Analyze latency, throughput, memory utilization, power consumption, and performance per watt.
- Profile workloads and identify bottlenecks across frameworks, runtimes, operating systems, memory subsystems, and hardware.
- Optimize workloads through quantization, graph optimization, kernel tuning, memory optimization, and heterogeneous compute utilization.
- Conduct competitive benchmarking across AMD and competing processor, GPU, APU, and AI accelerator platforms.
- Automate benchmark execution, data collection, regression analysis, and results visualization.
- Translate benchmarking results into recommendations for software, hardware, product, and business teams.
- Collaborate with architecture, framework, runtime, driver, firmware, and system software teams to resolve performance issues.
Requirements
- 5–10 years of relevant experience in AI performance engineering, system performance analysis, workload optimization, or benchmarking.
- Hands-on experience benchmarking AI inference workloads on CPU, GPU, APU, or NPU platforms.
- Strong experience with AI frameworks and runtimes including PyTorch, ONNX Runtime, vLLM, llama.cpp, or TensorFlow.
- Strong Linux expertise, including performance profiling, workload tracing, automation, and system-level debugging.
- Experience with profiling tools such as Linux perf, AMD uProf, ROCm Profiler, Nsight Systems, VTune, or Perfetto.
- Understanding of CPU/GPU architecture, memory hierarchy, power and thermal behavior, and performance trade-offs.
- Proficiency in Python and Bash for automation, data analysis, and reporting.
- Ability to communicate performance data through technical reports, dashboards, and recommendations.
- Bachelor’s or master’s degree in Computer Engineering, Electrical Engineering, Computer Science, Artificial Intelligence, Machine Learning, or a related field.
- Preferred: experience with ROCm, Ryzen AI Software, Vitis AI, MIGraphX, model quantization, mixed-precision inference, graph optimization, LLM serving, MLPerf Inference, Docker, virtualization, or post-silicon performance validation.
Benefits
- AMD benefits are referenced through its benefits-at-a-glance information.
Categories
About AMD
AMD designs and sells CPUs, GPUs, and adaptive/embedded computing products for PCs, data centers, gaming, and edge devices. Its portfolio includes Ryzen and EPYC processors, Radeon and Instinct graphics, and adaptive SoCs from its Xilinx acquisition, sold to OEMs, cloud providers, and device makers. Founded in 1969 and headquartered in Santa Clara, it is a public company on NASDAQ and supplies semi-custom chips for major game consoles.