Reproduce “Evaluating Computing Platforms for Sustainability: A Comparative Analysis of FPGAs against ASICs, GPUs, and CPUs” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Chetan Choppali Sudarshan, Aman Arora, Vidya A Chhabria
Arizona State University researchers developed GreenFPGA, a lifecycle-aware modeling framework that assesses the total carbon footprint of computing platforms including FPGAs, ASICs, GPUs, and CPUs. This tool integrates probabilistic modeling, device replacement effects, and reliability-aware degradation to offer a comprehensive comparison of sustainability across various deployment scenarios, identifying when FPGAs offer environmental advantages.
Key insights
- The explicit inclusion of End-of-Life (EOL) driven device replacement effects and memory CFP significantly impacts total carbon footprint estimations, with memory often contributing 50-70% of processor CFP.
- Probabilistic modeling provides a more nuanced understanding of sustainability outcomes, showing a range of probabilities for one platform to be more sustainable at specific points, rather than fixed deterministic crossover points.
- FPGAs can offer sustainability advantages, particularly with increasing numbers of applications due to embodied CFP amortization, and are generally more sustainable than CPUs in most scenarios due to power efficiency.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.