arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2511.04136cs.ETphysics.app-phphysics.optics
收藏

基于CMOS兼容平台的大规模光电神经元实现transformer-based LLMs

Implementation of transformer-based LLMs with large-scale optoelectronic neurons on a CMOS compatible platform

Artilux Inc.(阿特利克斯公司);Intel Corporation(英特尔公司);Department of Engineering, The University of Massachusetts, Boston(麻省大学波士顿分校工程学院)

Neil Na, Chih-Hao Cheng, Shou-Chen Hsu, Che-Fu Liang, Chung-Chih Lin, Nathaniel Y. Na, Andrew I. Shieh, Erik Chen, Haisheng Rong, Richard A. Soref

AI总结:

本文提出利用新型大规模光电神经元实现transformer模型,展示在40nm CMOS工艺下,GPT-3的1750亿参数在12.6 POPS速度下运行,具有高能效和高面积效率。

摘要:

近期数据中心基础设施的快速部署用于执行大型语言模型(LLMs)及相关人工智能(AI)应用,预计将在短期内导致能源消耗呈指数级增长。本文提出并分析了基于互补金属氧化物半导体(CMOS)兼容平台的transformer模型实现,该模型是现代LLMs的核心。通过将所有必需的光电设备和电子电路集成在仅约2厘米乘3厘米大小的芯片上,1750亿参数的GPT-3在仅使用40nm CMOS工艺节点的情况下,展示了前所未有的12.6 POPS推理速度,由无数据倾斜的光电版本张量阵列协调,同时具有74 TOPS/W的高能效和19 TOPS/mm2的高面积效率。对量化格式和硬件诱导误差的影响进行了数值研究,结果显示其影响极小。本研究为补充现有数字处理单元的模拟神经处理单元(NPUs)提供了一条新且实用的路径。

英文摘要:

The recent rapid deployment of datacenter infrastructures for performing large language models (LLMs) and related artificial intelligence (AI) applications in the clouds is predicted to incur an exponentially growing energy consumption in the near-term future. In this paper, we propose and analyze the implementation of the transformer model, which is the cornerstone of the modern LLMs, with novel large-scale optoelectronic neurons (OENs) constructed over a complementary metal-oxide-semiconductor (CMOS) compatible platform. With all of the required optoelectronic devices and electronic circuits integrated in a chiplet only about 2 cm by 3 cm in size, 175 billon parameters in the case of GPT-3 are shown to perform inference at an unprecedented speed of 12.6 POPS using only 40 nm CMOS process node, orchestrated by an optoelectronic version of systolic array with no data skew and negligible propagation delay, along with a high power efficiency of 74 TOPS/W and a high area efficiency of 19 TOPS/mm2. The influence of the quantization formats and the hardware induced errors are numerically investigated, and are shown to have a minimal impact. Our study presents a new yet practical path toward analog neural processing units (NPUs) to complement existing digital processing units.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑