Job TitleResponsibilities: Develop compute-intensive and highly performant test applications running on the MAIA hardware systems, focusing on PCIe, Memory, back-end network & TDP workloads. Design and develop workload profiling and analysis tools that capture and benchmark the runtimes of various compute blocks and/or hardware accelerators. Develop end-to-end post-silicon tests and tools for functional and performance scenarios of the system. Collaborate across organizations with AI Frameworks SW, Silicon & System Validation, Architecture and Firmware.Skill Sets: 8+ years' experience in software development with C++ and Python. 8+ years' experience in post-silicon SoC or system validation or diagnostic/stress test content development.
Experience with one or more of these: DMAs, PCIe, Memory (DDR, HBM), Network Controllers, Image decoders. Plus: Knowledge of or experience with AI models/kernels such as Gemm or Pytorch. Knowledge of power and thermal profiling, TDP/PnP, and PVT characterization.
Experience with Cuda or GPU or tensor-based programming is a plus.
Experience in build systems such as CMake and familiarity with CI/CD systems.