Databricks strikes 5-year deal to use Amazon AI chips - WSJ

Investing.com -- According to an article in the Wall Street Journal on Tuesday, Databricks and Amazon (NASDAQ:AMZN ) have entered a five-year partnership focused on using Amazon’s Trainium AI chips. The move is anticipated to reduce costs for businesses building their own AI tools.

The collaboration is said to allow Databricks to leverage Amazon’s custom AI chips to enhance its service for companies developing AI models.

The partnership comes as competition intensifies among tech firms like Microsoft (NASDAQ:MSFT ), Salesforce (NYSE:CRM ), and Snowflake (NYSE:SNOW ) for corporate AI dollars, said the WSJ.

Naveen Rao, vice president of generative AI at Databricks, highlighted the financial benefits of using Amazon’s Trainium chips.

The WSJ said the reduced costs could make AI tools such as bespoke customer service chatbots more accessible, helping companies save on staffing costs.

Databricks, valued at $43 billion as of last September, continues to expand its AI capabilities following its acquisition of AI startup MosaicML for $1.3 billion.

The company’s cloud-based software integrates analytics and AI, helping firms build tailored technology solutions.

While Amazon is not yet seen as a leader in AI innovation, the WSJ reported that the retail giant aims to position itself as a neutral AI provider through offerings like Trainium and Inferentia chips.

The chips, optimized for AI workloads, are designed to offer efficiency and cost savings, with Amazon claiming they are 40% cheaper than Nvidia’s GPUs.

The companies already have an existing partnership where Databricks runs on Amazon Web Services (AWS). Rao noted that customers using AWS have generated over $1 billion for Databricks, making AWS its fastest-growing cloud partner.

The deal underscores Amazon’s efforts to boost adoption of its AI chips amid competition from rivals like Nvidia (NASDAQ:NVDA ), AMD (NASDAQ:AMD ), and Google (NASDAQ:GOOGL ).

Source: Investing.com