2024 Intern
Architecture Modeling
Confirmed live in the last 24 hours
Samsung Research America

501-1,000 employees

The American arm of the R&D hub of Samsung Electronics
Company Overview
SRA is on a mission to create superior products and services that contribute to a better global society. SRA is the American arm of the R&D hub for Samsung Electronics.
Hardware

Company Stage

N/A

Total Funding

N/A

Founded

1988

Headquarters

Mountain View, California

Growth & Insights
Headcount

6 month growth

0%

1 year growth

17%

2 year growth

29%
Locations
Mountain View, CA, USA
Experience Level
Intern
Desired Skills
Python
CategoriesNew
Software Engineering
Requirements
  • Pursuing a Master of Science degree in Electrical/Computer Engineering or related discipline
  • Proficient in Python programming
  • Understanding of neural network and neural network accelerator IP
  • Familiarity with SOCs (System on Chip)
  • Experience with System-C based performance modeling or similar tools
  • Experience in creating or integrating simulation models of multi-core SoC subsystems
Responsibilities
  • Perform performance modeling and analysis of hardware features, applications, benchmarks, and uses cases
  • Analyze system-level architectural trade-offs and right size IPs within PPA constraint
  • Develop SoC performance models to evaluate PPA targets including CPU, GPU, ISP, NPU for mobile platforms
  • Leverage industry standard tools or create in-house tools for performance evaluation and analysis
  • Collaborate with teams to analyze workload, decompose and map workload components to IPs for system optimization
  • Work with various teams to identify, analyze, and optimize critical use cases including immersive, camera, AI, Display, Multimedia, and gaming applications
  • Evaluate architecture proposal performance benefits and communicate results across related engineering audiences
Desired Qualifications
  • C/C++ programming skills
  • Experience with Platform Architect, ModelSim, or VisualSim
  • Familiarity with cycle-accurate and TLM abstraction levels