Loading...
SeniorFull-timeAIML
Salary Range
$66,300 - $89,700 /year
This salary is estimated based on similar roles. The actual salary may vary.
About the Role
We are looking for an AI/ML ASIC Architect to develop advanced system architectures and AI/ML Accelerator ASIC architecture specifications for Sandisk’s next generation products. This role involves driving the frontend architecture of AI/ML Accelerator products and leveraging state-of-the-art interfaces like UCIe, PCIe, and CXL.
Key Responsibilities
- →Drive the AI/ML ASIC architecture that integrates AI Storage with GPU/TPU/xPU accelerators, focusing on I/O subsystems connected over UCIe/PCIe/CXL
- →Author architecture specifications for AI/ML xPU based Accelerator using AI Storage Solutions
- →Define I/O subsystem and PCIe DMA architectures, including interactions with internal embedded processor-subsystems, Network on Chip, Memory controllers, and FPGA fabric
- →Create flexible and modular I/O subsystem architectures deployable in Chiplet, monolithic, or 3D form factors
- →Work with customers and cross-functional teams to scope SoC requirements, analyze PPA tradeoffs, and define architectural requirements
- →Guide and assist pre-silicon design/verification and post-silicon validation
- →Improve AI/ML ASIC Architecture performance through hardware & software co-optimization
- →Analyze LLM Workload and competitive datacenter and AI solutions to identify performance improvement opportunities
- →Architect memory-efficient inference/training systems utilizing techniques like pruning, quantization, continuous batching, and speculative decoding
- →Collaborate with internal and external stakeholders/ML researchers
Requirements
- ✓Experience architecting components of AI/ML accelerator ASICs such as HBM, PCIe/UCIe/CXL, NoC, DMA, Firmware Interactions, NAND, xPU, fabrics, etc
- ✓Strong technical expertise in AI/ML systems and architectures
- ✓Excellent communication skills to collaborate with design and product planning teams
- ✓Ability to work with complex challenges and innovate in data centric architecture paradigm
Free to sign up. Takes 30 seconds.
