Google's Gemma is now optimized for NVIDIA's AI platforms
The teams at NVIDIA and Google have improved the performance of Gemma across NVIDIA GPUs in data centers and in RTX cards on local PCs.
As more companies look to improve their use of artificial intelligence, more tools and meshing between these companies is likely to happen. The latest case of this is between NVIDIA and Google, with the two working together to optimize Google’s Gemma across all NVIDIA AI platforms.
Announced on February 21, 2024 via the NVIDIA blog, Google’s Gemma (a 2 billion and 7 billion parameter open language model) has been optimized for use on all of NVIDIA’s AI platforms. This includes data centers, cloud computing, and even in private PCs using NVIDIA RTX GPUs.
The press release notes that developers will be able to run Gemma on cloud-based GPUs, including “Google Cloud’s A3 instances based on the H100 Tensor Core GPU and soon, NVIDIA’s H200 Tensor Core GPUs.”
Further support for Gemma will soon be available with Chat with RTX. This system lets users have access to generative AI capabilities on their own local PCs, provided they’re using an NVIDIA RTX GPU. This system effectively allows users to provide local files on their PC to a local AI in order to personalize the experience. NVIDIA notes that this creates fast results and because the user data stays on their PC, sensitive data isn’t shared with third parties.
This announcement comes alongside NVIDIA reporting its Q4 2024 earnings. The company managed to beat revenue and EPS expectations, with NVIDIA even setting Q1 2025 revenue guidance above Wall Street analyst expectations. Take a look at our NVIDIA and AI pages for more information.
-
Shacknews
replySam Chandler posted a new article, Google's Gemma is now optimized for NVIDIA's AI platforms
