Computing in the Era of Large Generative Models: From Cloud-Native to AI-Native

TL;DR

This paper explores the integration of large generative AI models with cloud-native architectures, proposing an AI-native paradigm that leverages cloud technologies and advanced ML runtimes to optimize costs and resource access.

Contribution

It introduces an AI-native computing paradigm combining cloud-native tech and ML runtimes, addressing challenges of large models like cost and resource efficiency.

Findings

Proposes an AI-native paradigm for large models

Highlights benefits of cloud-native integration

Encourages future research in AI-native computing

Abstract

In this paper, we investigate the intersection of large generative AI models and cloud-native computing architectures. Recent large models such as ChatGPT, while revolutionary in their capabilities, face challenges like escalating costs and demand for high-end GPUs. Drawing analogies between large-model-as-a-service (LMaaS) and cloud database-as-a-service (DBaaS), we describe an AI-native computing paradigm that harnesses the power of both cloud-native technologies (e.g., multi-tenancy and serverless computing) and advanced machine learning runtime (e.g., batched LoRA inference). These joint efforts aim to optimize costs-of-goods-sold (COGS) and improve resource accessibility. The journey of merging these two domains is just at the beginning and we hope to stimulate future research and development in this area.