AI-assisted development and validation for semiconductor design intelligence - ON-1265
Genre de projet: RechercheDiscipline(s) souhaitée(s): Génie - informatique / électrique, Génie, Informatique, Sciences mathématiques, Statistiques / études actuarielles
Entreprise: Advanced Micro Consulting Inc. (AMCI)
Durée du projet: 6 mois à 1 an
Date souhaitée de début: Dès que possible
Langue exigée: Anglais
Emplacement(s): ON, Canada
Nombre de postes: 2
Niveau de scolarité désiré: Études de premier cycle/baccalauréatMaîtrise
Ouvert aux candidatures de personnes inscrites à un établissement à l’extérieur du Canada: No
Au sujet de l’entreprise:
The proposed project will investigate how AI can help semiconductor engineers interpret design-report evidence while improving the reliability of the software and workflows used to deliver that intelligence. The initial focus is MachAI Lens, which compares timing results across builds and operating scenarios using reports from existing EDA tools.
With an academic supervisor and AMCI engineering guidance, the researcher(s) would define a focused use case, prototype selected AI-assisted functionality, and build repeatable validation methods. Proposed tasks include preparing permitted or synthetic report examples; comparing AI-generated explanations against source evidence and expert review; measuring incorrect or unsupported outputs; and developing regression checks for selected product and GUI workflows. A related option is evaluating AI-assisted prompting, code review and debugging against agreed quality and development-time baselines.
Expected outputs are a documented prototype or workflow improvement, a reproducible evaluation/test suite, measured results and recommendations for further development. Research questions include how to preserve traceability to evidence, evaluate usefulness to engineers and improve reliability across differing report formats and scenarios. Scope, research methodology, deliverables and program stream will be finalized with Mitacs and the academic partner. Only authorized, de-identified or synthetic data would be used, subject to confidentiality and IP arrangements.
Veuillez décrire le projet.:
The proposed project will investigate how AI can help semiconductor engineers interpret design-report evidence while improving the reliability of the software and workflows used to deliver that intelligence. The initial focus is MachAI Lens, which compares timing results across builds and operating scenarios using reports from existing EDA tools.
With an academic supervisor and AMCI engineering guidance, the researcher(s) would define a focused use case, prototype selected AI-assisted functionality, and build repeatable validation methods. Proposed tasks include preparing permitted or synthetic report examples; comparing AI-generated explanations against source evidence and expert review; measuring incorrect or unsupported outputs; and developing regression checks for selected product and GUI workflows. A related option is evaluating AI-assisted prompting, code review and debugging against agreed quality and development-time baselines.
Expected outputs are a documented prototype or workflow improvement, a reproducible evaluation/test suite, measured results and recommendations for further development. Research questions include how to preserve traceability to evidence, evaluate usefulness to engineers and improve reliability across differing report formats and scenarios. Scope, research methodology, deliverables and program stream will be finalized with Mitacs and the academic partner. Only authorized, de-identified or synthetic data would be used, subject to confidentiality and IP arrangements.
Expertise ou compétences exigées:
Proposed profile: computer science, electrical/computer engineering or a related discipline, with strong software development, testing and analytical skills. Relevant skills include Python or comparable programming, Git, structured/unstructured data processing, experimental design, evaluation metrics and technical documentation. Interest or experience in AI/LLMs, prompt evaluation and human-reviewed AI output is valuable.
Assets: semiconductor design, static timing analysis, Synopsys/Cadence report familiarity, automated GUI/regression testing and AI-assisted development tools such as Claude Code or ChatGPT. Prior EDA experience would be helpful but need not be mandatory for every role. Candidates should communicate clearly and respect confidential engineering data. Final requirements will follow the agreed project scope.
