Master Thesis: AI for Silicon Design
Om jobbet
Join our Team
About this opportunity:
Modern silicon design and System-on-Chip (SoC) development require running millions of EDA jobs weekly across diverse compute environments and vendor tools (Synopsys, Cadence, Siemens EDA, etc.). Our internal Run & Compile platform manages this complexity, supporting over 500 designers and verification engineers in flows like RTL linting (e.g., VC SpyGlass), simulation and verification (e.g., Xcelium, VCS), power and performance analysis, and regression and coverage. The platform executes hundreds of thousands of jobs daily, ensuring flow consistency, improving compute efficiency, and maintaining productivity.
The current orchestration gives scale and automation, but the next step is adding intelligence. To support advanced design methods, the platform must move beyond simple scheduling and become a self-optimizing, knowledge-aware, AI-augmented system that learns from past runs and improves outcomes. This change is important to meet aggressive time-to-market targets and to ensure Tape-Out quality and predictability in increasingly complex SoC projects.
This Master's thesis aims to build an AI-driven platform and reinforcement learning framework that enables:
Key Research Questions:
In this thesis, you will join the Silicon Methodology and Frontend CAD R&D team, collaborating with domain experts driving automation, verification, and design enablement. You'll gain hands-on experience with modern EDA infrastructure, HPC orchestration, and AI/ML applied to silicon development, contributing directly to how future SoCs reach Tape-Out.
What you will do:
Knowledge Assistant:
AI Infrastructure for Data Harvesting:
Reinforcement Learning for Flow Optimization:
The skills you bring:
What happens once you apply?
Click Here to find all you need to know about what our typical hiring process looks like.
We encourage you to consider applying to jobs where you might not meet all the criteria. We recognize that we all have transferrable skills, and we can support you with the skills that you need to develop.
Encouraging a diverse and inclusive organization is core to our values at Ericsson, that's why we champion it in everything we do. We truly believe that by collaborating with people with different experiences we drive innovation, which is essential for our future growth. We encourage people from all backgrounds to apply and realize their full potential as part of our Ericsson team. Ericsson is proud to be an Equal Opportunity Employer. learn more.
Primary country and city:Sweden (SE) || Stockholm
Primary Recruiter:Mia Erfält
About this opportunity:
Modern silicon design and System-on-Chip (SoC) development require running millions of EDA jobs weekly across diverse compute environments and vendor tools (Synopsys, Cadence, Siemens EDA, etc.). Our internal Run & Compile platform manages this complexity, supporting over 500 designers and verification engineers in flows like RTL linting (e.g., VC SpyGlass), simulation and verification (e.g., Xcelium, VCS), power and performance analysis, and regression and coverage. The platform executes hundreds of thousands of jobs daily, ensuring flow consistency, improving compute efficiency, and maintaining productivity.
The current orchestration gives scale and automation, but the next step is adding intelligence. To support advanced design methods, the platform must move beyond simple scheduling and become a self-optimizing, knowledge-aware, AI-augmented system that learns from past runs and improves outcomes. This change is important to meet aggressive time-to-market targets and to ensure Tape-Out quality and predictability in increasingly complex SoC projects.
This Master's thesis aims to build an AI-driven platform and reinforcement learning framework that enables:
- Knowledge Intelligence - converting operational data, logs, and user interactions into actionable insights and automated guidance for design and verification engineers.
- Adaptive Optimization - enabling the orchestration engine to learn from workload patterns to continuously enhance job performance, resource use, and flow health.
- Methodology Enablement - empowering the silicon methodology team to define, evolve, and validate next-generation flows using real-time intelligence and data-driven feedback loops.
Key Research Questions:
- How can knowledge-assisted AI improve productivity and decision-making in silicon design flows?
- What reinforcement learning strategies best balance speed, license constraints, and job success rate?
- How can orchestration data be transformed into a methodology intelligence layer for continuous flow evolution?
In this thesis, you will join the Silicon Methodology and Frontend CAD R&D team, collaborating with domain experts driving automation, verification, and design enablement. You'll gain hands-on experience with modern EDA infrastructure, HPC orchestration, and AI/ML applied to silicon development, contributing directly to how future SoCs reach Tape-Out.
What you will do:
Knowledge Assistant:
- Develop an intelligent assistant to answer flow-related questions, diagnose job failures, and explain EDA tool behaviors.
- Integrate NLP and retrieval-augmented generation (RAG) with documentation, logs, and historical data.
- Enable self-service troubleshooting and contextual design guidance.
AI Infrastructure for Data Harvesting:
- Build a scalable data pipeline to collect, curate, and correlate metadata from millions of jobs (runtime, logs, resource usage, success/failure metrics).
- Define feature representations for EDA workloads and flow topologies suited for ML and RL models.
- Ensure secure data governance and anonymization following enterprise policies.
Reinforcement Learning for Flow Optimization:
- Design RL agents to recommend or automatically tune runtime parameters (e.g., job partitioning, parallelization, memory tuning, license affinity).
- Optimize throughput, turnaround time, resource efficiency, and success rates.
- Implement feedback loops for continuous flow-level performance improvement.Ensure secure data governance and anonymization.
The skills you bring:
- Currently pursuing a MSc in Computer Science, Electrical Engineering, Engineering Physics, Embedded Systems, or related field.
- Strong interest in AI/Machine Learning, EDA systems, or Silicon Design Methodology.
- Proficiency in Python, data engineering, ML frameworks (PyTorch/TensorFlow), and familiarity with EDA flows are highly valued.
What happens once you apply?
Click Here to find all you need to know about what our typical hiring process looks like.
We encourage you to consider applying to jobs where you might not meet all the criteria. We recognize that we all have transferrable skills, and we can support you with the skills that you need to develop.
Encouraging a diverse and inclusive organization is core to our values at Ericsson, that's why we champion it in everything we do. We truly believe that by collaborating with people with different experiences we drive innovation, which is essential for our future growth. We encourage people from all backgrounds to apply and realize their full potential as part of our Ericsson team. Ericsson is proud to be an Equal Opportunity Employer. learn more.
Primary country and city:Sweden (SE) || Stockholm
Primary Recruiter:Mia Erfält
Ericsson AB
FöretagEricsson AB
Visa alla jobb för Ericsson AB