Nvidia Acquires Synthetic Data Firm Gretel To Train AI Models: Report

On Wednesday, Nvidia NVDA completed the acquisition of Gretel, a synthetic data startup.

According to a Wired report, the deal is valued over Gretel's recent valuation of $320 million. The exact financial terms of the purchase were not undisclosed.

The acquisition marks an expansion for Nvidia as it integrates Gretel's team and technology into its growing portfolio of cloud-based AI services, according to Wired.

Also Read: NVIDIA And Elon Musk's xAI Join BlackRock, Microsoft Led AI Infrastructure Partnership

Founded in 2020, Gretel is the generative AI platform for creating synthetic text, tabular, and time series data that includes tunable mathematically proven privacy and accuracy settings. Developers use Gretel to create artificial, privacy-enhanced versions of their sensitive data and to quickly generate new labeled samples to augment limited machine learning training datasets, all on-demand.

Gretel employs approximately 80 people and will enhance Nvidia's generative AI tools, particularly for developers creating AI models.

In December 2023, Gretel announced a partnership with Microsoft MSFT Azure and joined the Microsoft for Startups Pegasus Program.

The report mentioned that Nvidia has incorporated synthetic data tools into its product lineup, such as the Omniverse Replicator platform, which debuted in 2022.

On Wednesday, Nvidia launched several new products at the GTC event, including Blackwell Ultra, Rubin, and co-packaged optics technology that enables data centers to deploy more GPUs.

Nvidia has consistently achieved approximately a fourfold boost in AI computing power with each generation, from Hopper in 2022 to Blackwell in 2024, and the expected Rubin in 2026.

Additionally, NVIDIA unveiled a range of technologies aimed at advancing humanoid robot development, featuring the NVIDIA Isaac GR00T N1, the first fully open and customizable foundation model designed for general humanoid reasoning and skills.

Furthermore, Nvidia launched the Nemotron-4 340B mini-models, designed to produce synthetic data for multiple sectors.

Synthetic data created through computer simulations or generative AI models can help overcome these challenges. It includes text, videos, and images in both visual and non-visual formats, and can be combined with real-world data to train AI models more efficiently and affordably.

Price Action: NVDA shares traded lower by 0.06% at $117.45 in the premarket session at last check on Thursday.

Read Next: