AI Chip Architecture Engineer – Hardware/Software Co-Design
CANAAN CREATIVE GLOBAL PTE. LTD.
Singapore, Singapore
Not specified
Computer architecture knowledge
Transformer and llm workload analysis
Hardware-software co-design strategies
Canaan Creative Global Pte. Ltd. is seeking an AI Chip Architecture Engineer with expertise in hardware-software co-design, focusing on next-generation AI inference chips. The role involves defining and evaluating architectural features for AI accelerators, with a particular emphasis on Transformer and LLM workloads, and requires strong collaboration across various teams
Job Summary
The role involves defining and evaluating architectures for next-generation AI inference chips optimized for Transformer and LLM workloads.
Candidates will collaborate across algorithm, compiler, RTL, and software teams to deliver implementable hardware-software co-design solutions.
The position requires analyzing performance bottlenecks in power, area, memory, and latency while exploring AI accelerator microarchitectures.
Matching Summary
Match Score: 85
Canaan Creative Global Pte. Ltd. is seeking an AI Chip Architecture Engineer with expertise in hardware-software co-design, focusing on next-generation AI inference chips. The role involves defining and evaluating architectural features for AI accelerators, with a particular emphasis on Transformer and LLM workloads, and requires strong collaboration across various teams.
Skills & Requirements
Must-have
Computer architecture knowledge
Transformer and LLM workload analysis
Hardware-software co-design strategies
Memory hierarchy and interconnect design
Compiler frameworks like MLIR or TVM
Nice-to-have
Patent proposal experience
Strong analytical and communication skills
Fresh graduates with relevant projects
Novel architecture innovation ideas
Collaboration across algorithm teams
Key Requirements
Bachelor's degree in Electrical Engineering, Computer Engineering, or Computer Science
Experience with Verilog, SystemVerilog, VHDL, or Chisel
Familiarity with quantization, KV cache, prefill, and decoding components