Evaluation of pre-training large language models on leadership-class supercomputers

J Yin, S Dash, J Gounley, F Wang… - The Journal of …, 2023 - Springer
Large language models (LLMs) have arisen rapidly to the center stage of artificial
intelligence as the foundation models applicable to many downstream learning tasks …

A Comprehensive Performance Study of Large Language Models on Novel AI Accelerators

M Emani, S Foreman, V Sastry, Z **e, S Raskar… - arxiv preprint arxiv …, 2023 - arxiv.org
Artificial intelligence (AI) methods have become critical in scientific applications to help
accelerate scientific discovery. Large language models (LLMs) are being considered as a …

The case for co-designing model architectures with hardware

Q Anthony, J Hatef, D Narayanan, S Biderman… - Proceedings of the 53rd …, 2024 - dl.acm.org
While GPUs are responsible for training the vast majority of state-of-the-art deep learning
models, the implications of their architecture are often overlooked when designing new deep …

A comprehensive evaluation of novel AI accelerators for deep learning workloads

M Emani, Z **e, S Raskar, V Sastry… - 2022 IEEE/ACM …, 2022 - ieeexplore.ieee.org
Scientific applications are increasingly adopting Artificial Intelligence (AI) techniques to
advance science. High-performance computing centers are evaluating emerging novel …

Comparative Study of Large Language Model Architectures on Frontier

J Yin, A Bose, G Cong, I Lyngaas, Q Anthony - arxiv preprint arxiv …, 2024 - arxiv.org
Large language models (LLMs) have garnered significant attention in both the AI community
and beyond. Among these, the Generative Pre-trained Transformer (GPT) has emerged as …

A scalable real-time data assimilation framework for predicting turbulent atmosphere dynamics

J Yin, S Liang, S Liu, F Bao, HG Chipilski… - SC24-W: Workshops …, 2024 - ieeexplore.ieee.org
AI-based foundation models like FourCastNet, GraphCast, ClimaX, and Pangu-Weather are
revolutionizing weather and climate predictions but are not yet ready for operational use …

Toward a holistic performance evaluation of large language models across diverse ai accelerators

M Emani, S Foreman, V Sastry, Z **e… - 2024 IEEE …, 2024 - ieeexplore.ieee.org
Artificial intelligence (AI) methods have become critical in scientific applications to help
accelerate scientific discovery. Large language models (LLMs) are being considered a …

Torchbench: Benchmarking pytorch with high api surface coverage

Y Hao, X Zhao, B Bao, D Berard, W Constable… - arxiv preprint arxiv …, 2023 - arxiv.org
Deep learning (DL) has been a revolutionary technique in various domains. To facilitate the
model development and deployment, many deep learning frameworks are proposed, among …

Stencil Computations on AMD and Nvidia Graphics Processors: Performance and Tuning Strategies

J Pekkilä, O Lappi, F Robertsén… - arxiv preprint arxiv …, 2024 - arxiv.org
Over the last ten years, graphics processors have become the de facto accelerator for data-
parallel tasks in various branches of high-performance computing, including machine …

LLM-Inference-Bench: Inference Benchmarking of Large Language Models on AI Accelerators

KT Chitty-Venkata, S Raskar, B Kale… - SC24-W: Workshops …, 2024 - ieeexplore.ieee.org
Large Language Models (LLMs) have propelled groundbreaking advancements across
several domains and are commonly used for text generation applications. However, the …