ML Accelerator Performance Validator – Post‑Silicon

Amazon; Austin, Northern (TX, KY); Hybrid; USD 144,000 - 194,000; Full time

Job summary

Amazon’s Annapurna Labs in Austin, TX seeks an ML Accelerator Performance Validation Engineer to quantify and qualify the performance of AWS's custom ML training chips against architectural targets. You will bridge silicon capabilities with real-world ML workloads to ensure latency, throughput, and efficiency meet cloud-scale demands.

You will design benchmarks, profile workloads, identify bottlenecks, and build automated dashboards.

Qualifications

• 3+ years of non‑internship professional software development experience.
• 2+ years of non‑internship design or architecture experience.
• Experience with ML and LLM fundamentals, or PyTorch/JAX.
• Bachelor’s degree in CS/Engineering/Math or equivalent, or experience in Java, C++, Python, or related language.
• 3+ years of experience with hardware performance counters and profiling tools.
• Strong understanding of computer architecture fundamentals (memory hierarchies, compute pipelines, interconnects).
• Experience with statistical methods, regression analysis, and data visualization to interpret performance data.
• 3+ years of full software development life cycle experience.
• Experience with CUDA kernels or ML low-level kernels or deploying LLMs on GPUs/accelerators.
• Knowledge of collective communications and scaling.
• Experience with HBM, PCIe, DMA bandwidth characterization.

Responsibilities

• Design and execute performance benchmarks spanning micro-architectures to full model training.
• Measure and analyze compute throughput, memory bandwidth, interconnect latency, and more.
• Profile real ML workloads (transformer models, LLMs, vision models) on silicon.
• Identify performance bottlenecks and work with architecture teams on optimization.
• Build automated performance regression dashboards and tracking infrastructure.
• Correlate silicon measurements against RTL simulation and emulation predictions.

Skills

Machine Learning fundamentals

Large Language Models

CUDA kernels

Java/C++/Python

Hardware performance profiling

Computer architecture

Data visualization

Software development lifecycle

LLM deployment on AI accelerators

Collective communications (AllReduce/​

Education

Bachelor's degree in CS/Engineering/Math or equivalent

Tools

PyTorch

JAX

Job description

Amazon’s Annapurna Labs in Austin, TX seeks an ML Accelerator Performance Validation Engineer to quantify and qualify the performance of AWS's custom ML training chips against architectural targets. You will bridge silicon capabilities with real-world ML workloads to ensure latency, throughput, and efficiency meet cloud-scale demands.

You will design benchmarks, profile workloads, identify bottlenecks, and build automated dashboards.