• Limitations of Simple Additive and Multiplicative Scoring Methods

  A critical review of Simple Additive Scoring (SAS) and Simple Multiplicative Scoring (SMS) in education, healthcare, GIS, and human development, showing where simple aggregation creates dangerous logic failures.
• Analyzing Performance in Stencil Operations using the Sobel Filter

  A GPU vs. CPU performance study of the Sobel edge detection algorithm using OpenMP, CUDA, and OpenMP GPU offload on Perlmutter's A100 nodes, with CUDA achieving 91.4% occupancy at optimal configuration.
• Comparing LIKWID Hardware Metrics to Performance across DGEMM Implementations

  Using the LIKWID performance monitoring tool to correlate L2/L3 cache accesses and retired instruction counts with DGEMM runtime, revealing that naive implementations can incur up to 600x more L3 accesses than CBLAS.

• Feb2025 - Current
  San Francisco State University
  Graduate Research Assistant / Lead Developer

  Lead developer for AMAVA, an accessibility research project translating video input into real-time audio feedback for visually impaired users.

  • Engineered a motion-aware audio engine using YOLOv2 to classify environmental context and dynamically throttle audio playback.
  • Built a low-latency pipeline with captioning and caching using Gemini, torchaudio, ElevenLabs, and FFmpeg.
  • Authored an 8-page research paper detailing system architecture, user trials, and performance benchmarks.
• Jan2025 - Current
  San Francisco State University
  Graduate Certificate in Software Engineering

  Focusing on AI, soft computing, high performance computing, and advanced software engineering principles. Expected completion 2026.

Let's Connect

• linkedin /
• benmklein@gmail.com