Overview In this role you define next gen cloud and datacenter processors by building high speed cycle accurate simulators and evaluating design trade offs. You will partner with CPU/Compute, SoC, and RTL teams to ensure top performance and power efficiency for hyperscale workloads. You'll address bottlenecks and drive architectural improvements with data driven analysis. This position offers the chance to shape silicon that powers modern data centers and AI/data processing pipelines.
Compensation / Benefits- comprehensive health insurance
- life and disability insurance
- savings plan
- Company paid holidays
- Sick Leave
- Parental leave
Responsibilities- Design, develop, and maintain cycle accurate and transaction level performance models for multi core datacenter SoCs (C++/SystemC).
- Analyze cloud-native workloads and AI/data pipelines to derive architectural requirements.
- Perform what if analyses and design sweeps to optimize cache, memory bandwidth, latency, interconnects, and core counts.
- Identify bottlenecks and propose hardware or software solutions.
- Correlate models with RTL, emulators, and silicon to ensure accuracy.
- Build tracing tools, dashboards, and automated regression pipelines using Python.
Key requirements- 5+ years in CPU/compute core/SoC/system level performance modeling and architectural analysis.
- Experience profiling hyperscaler workloads (SPEC CPU, cloud microservices, databases, virtualization).
- Expert-level modern C++ (C+/17+) and object oriented design; strong Python or Perl scripting.
- Deep knowledge of high performance CPU microarchitecture, memory hierarchies, cache coherency, and main memory tech (DDR, LPDDR5/6, HBM).
- Familiarity with on chip interconnects (AMBA CHI, AXI), PCIe/CXL, and multi die (UCIe).
- Experience with simulators (gem5, Sniper) or building trace driven/execution driven simulators; strong communication skills.
- Clear communicator
- Cross functional collaboration
- Analytical thinker
- C++ (14/17+)
- Python/Perl scripting
- SystemC
