Aramco Digital Company and Groq, a major name in the field of artificial intelligence (AI) inference, have announced a partnership to establish the world's largest inferencing data centre in Saudi Arabia. But it's far from the only recent AI news from the digital and technology subsidiary of Aramco, the giant global integrated energy and chemicals company.
AI inference, according to tech giant IBM, is the process of running live data through a trained AI model to make a prediction or solve a task. The task could be, for example, identifying spam, converting speech to text, or distilling a long document into key takeaways. The goal of AI inference is to calculate and output an actionable result.
Referring to the shift in AI towards deploying or running models, Groq says it can provide developers and enterprises with fast AI inference in the cloud and in on-premises AI compute centres.
Which is no doubt why the Saudi Press Agency describe this strategic collaboration as marking “a significant step forward in advancing the Kingdom's digital transformation initiatives and solidifying its position as a global leader in AI and cloud computing”.
There appear to be no specific details just yet about the building or equipping of the inferencing data centre. However, we are told that the facility will process 5.28 million tokens per second by the end of 2024 and 53 million tokens per second by 2025.
Inferencing speeds are measured in latency, or the time it takes for an AI model to generate a token — a word or part of word— when prompted.
As the Data Centre Dynamics website points out, the partnership with Groq will also help Aramco Digital bring its voice-commanded generative AI model Norous to market, with Groq using its AI semiconductor skills to provide the necessary scalable infrastructure.
Indeed, the website notes that Aramco Digital has announced a number of recent partnerships with companies operating in the AI and wireless technologies space.
For example, a collaboration with AI company Cerebras Systems will see the two companies aim to bring high-performance AI inference to industries, universities, and enterprises in Saudi Arabia. Under the terms of the agreement, Aramco will deploy Cerebras CS-3 systems into its cloud computing business to accelerate the building, training, and deployment of large language models (LLMs) and AI applications.
Aramco Digital and Qualcomm, meanwhile, are working with Saudi Arabia’s Research, Development and Innovation Authority (RDIA) to launch a startup incubator programme to support AI, wireless technologies, and IoT startups in the country, providing them with technical assistance, business coaching, and IP training.
Aramco Digital has also unveiled the world's first processors with native support for 5G in 450MHz spectrum, developed in partnership with wireless technology multinational Qualcomm.
