Summer 2025

2025 Intern

GPU Architecture Modeling

Posted on 2/27/2025

Samsung Research America

Samsung Research America

501-1,000 employees

Develops advanced technology in multiple domains

Compensation Overview

$48 - $63/hr

+ Bonus

Mountain View, CA, USA

Category
Embedded Systems Engineering
Electrical Engineering
Required Skills
OpenGL
C/C++
Requirements
  • Pursuing PhD degree in Electrical/Computer Engineering preferably specializing in computer graphics, or related discipline. Open to new grads however, to qualify for intern program, the internship must be completed within 1 year after official graduation
  • Proficient with modern Graphics processor architecture, 3D rendering pipelines & optimization and modern graphics APIs (OpenGL, Vulkan, DirectX etc)
  • Good Understanding of mapping AI models on GPU HW
  • Proficient in C/C++ and understanding of architecture model development using SystemC
  • System-C based performance modeling or some experience with Platform Architect
  • Good understanding of GPU & computer architecture
  • Experience creating or integrating simulation models of multi-core SoC subsystems at different levels of abstraction (e.g., cycle-accurate and TLM)
Responsibilities
  • Build GPU architectural models in SystemC/platform architect of existing and new GPU architectures.
  • Perform performance modeling and analysis of various GPU hardware features, applications, benchmarks, and uses cases.
  • Analyze system-level architectural trade-offs (across mapping of workload to CPU, GPU, NPU, ISP, memory subsystems, and system software). Right size GPU blocks within PPA constraints.
  • Develop SoC performance models to help evaluate PPA (power, performance, area) targets including CPU, GPU, ISP, NPU – particularly around mobile platforms.
  • Leverage Industry standard tools and/or create in house tools to speed up performance evaluation and analysis.
  • Work across the organization to analyze workload and decompose, map workload components to IPs to optimize system performance, energy.
  • Evaluate architecture proposal performance benefits in collaboration with SoC Architects and communicate the results across related engineering audiences (SW, HW, Architecture, Leadership).
Samsung Research America

Samsung Research America

View

Samsung Research America specializes in advanced technologies such as next-generation communications, artificial intelligence, digital media, mobile platforms, and digital health. Their innovations include FadeNet, a convolutional neural-network based technology, and the ECG Monitor App, which enables ECG recording using Galaxy Watch for atrial fibrillation screening.

Company Size

501-1,000

Company Stage

N/A

Total Funding

N/A

Headquarters

Mountain View, California

Founded

1988

Benefits

Onsite Fitness Facility

Subsidized Meals

Coffee Barista Bar

Samsung University

Product Discounts

Dog Friendly Campus

Wellness Programs

Activity Clubs

Growth & Insights

Headcount

6 month growth

↑ 20%

1 year growth

↑ 0%

2 year growth

↑ 20%
INACTIVE