Software Engineer - Hardware Abstraction Layer, AWS Machine Learning Accelerators
Core
Develop the hardware abstraction layer (HAL) for AWS custom machine learning system-on-chips (SoCs) like Trainium and Inferentia, managing low-level hardware details for cloud-scale ML infrastructure.
Role type
Senior IC software engineer (hardware abstraction layer)
Builds
HALs for newly developed SoC IPs and the underlying AWS infrastructure management software stack
Domain
Cloud infrastructure + custom silicon (ASIC/SoC) architecture
Deliverable
production ML models | infrastructure
Required skills
C++, Python, SoC/CPU/GPU/ASIC architecture, modular driver architectures, functional/performance modeling of hardware
Preferred skills
DV team integration, SystemC/TLM modeling, RTL simulation correlation, performance model calibration, Modern C++ (20+), PyTest/GoogleTest, multi-threaded simulation
Technologies
C++, Python, SystemC, TLM, PyTest, GoogleTest
Responsibilities
Build HALs for new SoC IPs, solve SoC/system-level architectural issues, test and deploy software stacks to internal customers, refactor and maintain codebases, innovate on customer tooling
Seniority
Senior, hands-on IC