AI/ML ASIC Architect

Responsibilities

• Drive AI/ML ASIC architecture integrating AI Storage with GPU/TPU/xPU accelerators via UCIe, PCIe, and CXL
• Author detailed architecture specifications for AI/ML xPU based accelerators
• Define I/O subsystem and PCIe DMA architectures, including interactions with embedded processors and Network on Chip
• Create modular I/O subsystem architectures for Chiplet, monolithic, or 3D form factors
• Collaborate with cross-functional teams to scope SoC requirements and analyze PPA tradeoffs
• Perform LLM workload analysis and characterization to identify performance improvement opportunities
• Architect memory-efficient inference/training systems using pruning, quantization, and speculative decoding

Requirements

• Bachelors with 7+ years, Masters with 5+ years, or PhD with 3+ years in Computer/Electrical Engineering
• 2+ years of experience in architecture authoring specifications
• Strong background in architecting ASIC, SoC, or I/O subsystems (PCIe/UCIe/CXL/DMA)
• Knowledge of I/O Subsystem and DMA interactions with x86, RISC-V, or ARM processors
• Understanding of computer/graphics architecture, ML, and LLM
• Experience with GPU/TPU/xPU Accelerator systems and high bandwidth memory hierarchy
• Proficiency in microarchitecture, software, and hardware performance engineering
• Knowledge of ARM Processors and AXI Interconnects

Preferred Qualifications

• Familiarity with UCIe, CXL, NVLink, or UAL protocols
• Experience with high-speed networking like InfiniBand or RDMA
• Expert knowledge of transformer architectures and attention mechanisms
• Expertise in CUDA programming and GPU memory hierarchies
• Experience with NVMe storage systems, protocols, and NAND flash

About the Company

Sandisk relentlessly innovates to deliver solutions that enable today’s needs and tomorrow’s next big ideas, specializing in Flash and advanced memory technologies.