Recent research from Omdia has revealed that major Cloud providers are in a competitive race to deliver AI at scale. Google Cloud Platform (GCP) emerges as the frontrunner in pioneering this, while Amazon Web Services (AWS) is dedicated to providing cost-effective solutions.
GCP benefits from Google’s status as a powerhouse in fundamental AI research, whereas AWS leverages both the vast scale of its existing business and its proficiency in day-to-day operations. Customers seeking to embrace cutting-edge technology are best served by GCP, while those prioritising affordability will find AWS most suitable. However, Microsoft Azure appears to be directing its efforts towards meeting OpenAI’s demand for capacity.
The research, recently published in Omdia’s report, “AI Inference Products & Strategies of the Hyperscale Cloud Providers,” explores how major Cloud infrastructure vendors serve inference – referring to the process of generating content or answers from an AI model once its training is concluded. Inference becomes essential when an AI application transitions into production, driven by end-user requirements, thereby representing the convergence of AI projects and practical implementation. As growing numbers of AI applications go into production, Omdia foresees inference occupying an increasingly significant portion of overall AI computing demand.
“The competition in this sector is intense. Google has an edge related to its strength in fundamental AI research, while AWS excels in operational efficiency, but both players have impressive custom silicon,” said Alexander Harrowell, Omdia’s Principal Analyst for Advanced Computing. “Microsoft took a very different path by concentrating on FPGAs initially but is now pivoting urgently to custom chips. However, both Google and Microsoft are considerably behind AWS in CPU inference. AWS’ Graviton 3 and 4 chips were clearly designed to offer a strong AI inference option and staying on a CPU-focused approach is advantageous for simplifying projects.”
Hyperscalers are providers of computing services to the majority of the AI industry, likely serving as the initial contact point for those establishing an AI model inference infrastructure to cater to users. Omdia’s report is designed to inform enterprises on crucial recommendations when selecting an appropriate provider and the array of options available. The study provides information on the pricing and availability of custom AI silicon, including Google TPUs, flagship, mid-range, and entry-level GPUs, as well as CPU options recommended by hyperscalers for AI inference.
There’s plenty of other editorial on our sister site, Electronic Specifier! Or you can always join in the conversation by commenting below or visiting our LinkedIn page.
