Oracle Embeds ChatGPT and Codex in Apps, Adds Google and OpenAI Chips to Cloud

Oracle announced on Tuesday that it is integrating OpenAI's ChatGPT and Codex tools into its software applications to enhance database functionality. The company also expanded its cloud hardware options by adding Google's tensor processing units and OpenAI's Jalapeño chip to its infrastructure lineup.

Mahesh Thiagarajan, Oracle's president of Cloud Infrastructure, spoke at the Oracle AI World conference in Las Vegas. He stated that the integration of ChatGPT allows users to build, deploy, and manage applications with greater ease. Thiagarajan noted that his team observed the OpenAI GPT-6 Astra model being trained at Oracle's Stargate data center campus in Abilene, Texas. He said the team could see real-time benefits from that process.

OpenAI CEO Sam Altman is scheduled to appear on stage with Oracle co-CEO Clay Magouyrk for a keynote address at 12:40 p.m. ET on Tuesday. Greg Brockman, president and co-founder of OpenAI, said in a press release that more businesses are choosing OpenAI because the company makes it easier to deploy AI within existing tools and systems.

The announcement marks a continued push by OpenAI into the enterprise sector. It occurred one day after Anthropic filed its S-1 document in preparation for an initial public offering. Oracle's database systems are widely used in healthcare and finance for logging and securing data.

Oracle also confirmed that Cerebras hardware will run on its cloud infrastructure. A Cerebras spokesperson said the deal is material, but the company will not disclose the scale because it is currently in a quiet period. Following the news, Cerebras stock rose by approximately 8 percent. Oracle shares gained nearly 5 percent on Tuesday.

Thiagarajan addressed the availability of different hardware chips. He said customers now have infinite possibilities for driving AI. When asked if this signaled a shift away from Nvidia GPUs, he said no. He described Nvidia chips as the best and most fungible, noting that the AI market remains incredibly large. He added that chip scarcity remains an ongoing issue, so having different accelerators available benefits customers like OpenAI.

This is the first time Google's custom chips will run on another hyperscaler's cloud. The arrangement shows how tech companies are seeking to place their hardware in the hands of frontier AI labs. This follows a similar move by Amazon Web Services, which announced plans seven months ago to bring Cerebras hardware into its data centers.

Thiagarajan also discussed Oracle's Project Jupiter, a multibillion dollar data center in New Mexico. He said the project is just one of 30 to 40 data center projects in development per quarter. He stated that Oracle is not too worried about local pushback because it is building global infrastructure. A hearing on air quality permitting is scheduled for December 7 in Sunland Park. Oracle maintained that Project Jupiter is on schedule after sending a force majeure notice to backer Blue Owl Capital last month. Oracle shares are down more than 30 percent this year due to concerns over financing its digital infrastructure pipeline.