Principal Enterprise Performance Engineer: CPU & Database

AMD; Austin (TX); On-site; USD 150,000 - 190,000; Full time

Job summary

AMD is seeking an end-to-end performance engineer to own enterprise and database workloads, focusing on CPU behavior, memory hierarchy, and software tuning. You will be the primary workload authority for AMD EPYC versus Intel Xeon and evaluate Arm platforms for enterprise use cases.

You will diagnose results, separate CPU limits from storage/network/runtime effects, and build reproducible methods for benchmarking across AMD platforms and partner environments.

Qualifications

• Hands-on performance engineering experience at scale with enterprise servers.
• Experience with relational/distributed databases and JVM performance.
• Strong Linux and system-administration skills; scripting and repeatable env setup.
• Proficiency with profiling tools and hardware counters; NUMA analysis.

Responsibilities

• Define and maintain an enterprise workload taxonomy across databases, Java, analytics, virtualization, and more.
• Develop rigorous methods for throughput, latency, scale-out, and consolidation.
• Analyze AMD vs Intel performance focusing on core config, frequency, power, memory, and platform setup.
• Evaluate tuning of compiler, JVM, databases, kernels to reflect realistic deployments.
• Use profiling to explain stalls, memory behavior, locks, and bottlenecks.
• Create deep-dive reports and executive content on lead/risk areas and mitigations.
• Support post-launch analysis to improve forecasts.

Skills

Performance engineering

Databases

Linux scripting

Profiling tools

NUMA analysis

System administration

Benchmark fairness

Education

Bachelor's or Master's in Electrical Engin

Tools

Oracle

SQL Server

PostgreSQL

MySQL

SAP

Java application servers

Analytics platforms

Virtualization stacks

Job description

AMD is seeking an end-to-end performance engineer to own enterprise and database workloads, focusing on CPU behavior, memory hierarchy, and software tuning. You will be the primary workload authority for AMD EPYC versus Intel Xeon and evaluate Arm platforms for enterprise use cases.

You will diagnose results, separate CPU limits from storage/network/runtime effects, and build reproducible methods for benchmarking across AMD platforms and partner environments.