IT

Chiplet Connectivity & SystemC/TLM Engineer

Accepting applications

Iravan Technologies · Pune City, Maharashtra, India

Full-Time Entry AIC++EthernetPCIeSerDes
Estimated market salary
₹7-12 LPA

This is a SiliconBoard market estimate, not an employer-posted salary.

Posted
3d ago
Category
Eda
Experience
Entry
Country
India
Designation: Chiplet Connectivity & SystemC/TLM Engineer
Experience Level: 5–10+ Years
Role Type: Individual Contributor
Location: Bangalore / Hyderabad (Hybrid)
Position Summary
Develop functional and performance models for the company’s chiplet connectivity and high-speed data-movement infrastructure, including UCIe, PCIe, CXL, and die-to-die interfaces.
Key Responsibilities
Connectivity Virtual Modelling: Develop SystemC/TLM models of connectivity components; model transactions, data movement, latency, and bandwidth; develop configurable models for different link configurations; represent software-visible behavior where required.
AI Scale-Up / Fabric Modelling: Model communication between multiple XPUs/chiplets, develop scalable interconnect/fabric models, and analyze bandwidth, latency, congestion, and utilization to support scale-up architecture exploration.
Multi-Chiplet / Multi-XPU Scalability: Integrate connectivity models into multi-chiplet configurations and evaluate communication scalability as the number of XPUs/chiplets increases.
Required Technical Skills
5–10+ years of relevant SystemC/TLM or connectivity-modelling experience.
Strong C++ / SystemC / TLM.
UCIe, PCIe, and CXL.
Chiplet architecture and high-speed interconnects.
Transaction-level modelling and performance modelling.
Preferred Skills
AXI / ACE / CHI.
Ethernet, UALink, ESUN, SerDes.
AI/HPC architecture.
Synopsys Virtualizer.
Education
B.E. / B.Tech / M.E. / M.Tech in Electronics, Electrical, Computer Engineering, or related discipline.
Project Experience
Hands-on development of SystemC/TLM connectivity or interconnect models (UCIe/PCIe/CXL or die-to-die) used for bandwidth/latency analysis.
Soft Skills
Strong modelling discipline and attention to protocol-level detail.
Ability to work closely with the platform architect and performance architect to integrate models cleanly.
Good to Have
Experience modelling scale-up fabrics for multi-XPU AI systems.
Familiarity with SerDes/UALink for next-generation interconnects.
Expected Deliverables
SystemC/TLM functional and performance models for UCIe/PCIe/CXL and die-to-die interfaces.
Scalable interconnect/fabric models integrated into multi-chiplet configurations.
Bandwidth/latency/congestion analysis reports.

Interested candidates may share resumes at [email protected](Immediate to 30-day notice period candidates preferred).
Show more Show less