Samba Nova: We aim to develop an advanced artificial intelligence system for Aramco at the lowest possible cost

Samba Nova Artificial Intelligence Company aims to develop an advanced artificial intelligence system for Saudi Aramco, at the lowest possible cost, to be used in its giant model “Aramco Meta Brain,” Marshall Choi, the company’s vice president of products, told Al-Eqtisadiah.

Aramco MetaBrain is a pioneering, large-scale, industrial-grade generative AI model, announced by Aramco President and CEO Amin Nasser at the third edition of the LEAP 2024 conference in Riyadh last March.

The model is based on cumulative data collected by the company over 90 years, and represents a strategic investment in generative AI by the company, which plans to use it to power cognitive applications across its business.

“We started working with Aramco about a year ago, and as a result of this cooperation, the company will benefit from the artificial intelligence we provide and employ it in all its operations, which will contribute to greatly enhancing the company’s productivity. Our mission is to provide a very advanced system at the lowest possible cost,” Choi said.

According to Choi, the system combines high-speed data analysis with accuracy in results, enabling developers to leverage the power of large, custom language models in their organizations to streamline workflows and accelerate innovation.

At the infrastructure level, he explained that the company provides its customers in general with the fastest devices, including chips, electronic chips, and systems, for training and reasoning purposes.

Regarding software, he pointed out that pre-trained generative AI models with more than a trillion parameters are provided based on open sources to enhance leadership capabilities in AI models at the regional level, relying on experience working in Silicon Valley.

This is in addition to establishing an AI infrastructure at the national level of countries or the level of large companies that helps transfer knowledge and skills related to AI to the local market, thus helping countries have their own path regarding AI, which enables them to maintain the privacy of sensitive data and preserve intellectual property rights. Graphics processing units have played an important role in the advancement of AI, providing unprecedented computing speed and performance, in addition to enabling amazing new commercial, scientific and governmental applications and capabilities, according to Samba Nova’s Vice President of Products.

However, the increasing prevalence of AI globally is driving demand for capabilities that require larger language models (LLMs), which require larger numbers of advanced GPUs, but the problem is that these units are expensive, difficult to obtain, very energy-intensive, and difficult to manage.

To overcome this problem, Marshall Choi said the company has launched a revolutionary chip called SN40L that dramatically reduces the number of processors needed to train enterprise AI models and achieves faster inference speeds.

He noted that these chips have great capabilities that enable them to train large language models and infer from them at tremendous speed, and provide solutions provided by competing companies that need separate systems (and thus more graphics processing units) to perform these tasks, and this leads to a large improvement in the use of devices, memory and energy, and reduces the total cost by up to 10 times.

In a related context, Khaled Al Shami, Vice President of Solutions Consulting at Infor, believes that countries, companies and institutions must develop strategies that are consistent with their goals and needs to benefit from artificial intelligence in improving production, services and developing products.

He pointed out that research and innovation in the field of artificial intelligence should be supported by enhancing cooperation between universities and companies to develop new and innovative technologies and solutions that use artificial intelligence. He added that research contributes to automation and improving productivity, as the use of smart technologies such as robots and smart systems in companies leads to increased productivity rates, improved quality of products and services, and improved efficiency in the use of resources.

He pointed out that the use of generative artificial intelligence models in many cases within companies and institutions has taken many forms, including writing emails, preparing reports, product documents and content on the web, in addition to creating job cards and job applications, comparisons between products and suppliers, and collecting images, music and video clips for marketing campaigns.