GPU Driver & Memory Management Architect
Core
Architect driver-level memory management and enable dynamic memory allocation across CPU, GPU, and NPU to optimize performance for AI and graphics hybrid workloads.
Role type
Senior IC GPU Driver & Memory Management Architect
Builds
GPU/NPU driver architectures and shared memory models for AI workloads
Domain
Semiconductor / Graphics / AI Infrastructure
Deliverable
production ML models
Required skills
Memory management, GPU architecture, OS-driver contracts, Performance optimization, Hardware-software integration, Debugging, FPGA, Systems design
Preferred skills
Electrical engineering, Hardware design, MATLAB, Schematic capture, Simulations
Technologies
WDDM, DVMT, Windows, Linux, FPGA, Oscilloscope
Responsibilities
Architect driver-level memory management (DVMT, shared memory models, WDDM), Enable dynamic memory allocation across GPU, CPU, and NPU, Optimize iGPU performance for AI + graphics hybrid workloads, Define OS-driver contracts (Windows/Linux), Diagnose and resolve memory-related performance bottlenecks, Partner with IHVs (Intel/AMD/Qualcomm) on driver capabilities and requirements
Seniority
Senior, hands-on IC
