Software System Design Eng. / India / Hyderabad / 6 Years Experience

Windows OS PNP power and performance Server server power Benchmarking SoC architecture setting up servers storage debug including networking Python

Server Power and Performance Optimization team is looking for a hands-on engineer to optimize the performance of important server workloads/benchmarks throughout the life of our server platforms.
As new silicon shows up, you will be trying to understand the performance of the silicon as well as the platform.
You will need to understand pre-silicon projections and how they map to reality. As the parts become generally available, you will be the expert who helps our customers (major OEMs, HPC, mega datacenters etc.) make their workloads run faster. Finally, you will analyze the competition and identify comparative strengths and weaknesses.
Candidate must have strong software chops and be capable of developing expertise in the AMD EPYC architectures.
Strong communications skills are required as this person will work with hardware and software architects across the company to identify potential improvements, characterize behavior and drive improvements wherever they lie.
Potential improvements could be in (but not limited to) the platform configuration, the operating system, the compiler, the BIOS, firmware or processor architecture. Candidate must be capable of studying and understanding complex workloads.

Requirements

6+ years of industry experience
Strong engineering background ranging from application level to BIOS
Must be able to understand detailed SoC architecture including peripherals and subcomponents
Experience with protocol and bus analyzers
Works with platform and software architects to understand system performance dials
Strong communications skills. Able to summarize complex problems for executives as well as drill down to details with architects and engineers utilizing low level debugging skills.
Strong analytic skills including the ability to analyze current behavior, identify potential areas for improvement and design / execute experiments with the overall goal being to optimize the performance of the workload
Comfortable setting up servers, changing hardware and installing server SW
Must be a self-starter and self-motivated.
Experience with system and software performance debug tools
Strong systems level debug including networking, storage, compute, clusters etc.

Bonus points:

Strong understanding of AMD EPYC architecture
Strong understanding of trade-offs between power and performance
Familiarity with the strengths and weaknesses of the NUMA architecture
Experience designing micro-architecture experiments to measure power and performance