NVIDIA supports enterprise-specific generative AI acceleration with 'NVIDIA AI Ready Servers'.

NVIDIA announced on the 24th that major system manufacturers around the world are providing NVIDIA AI-ready servers that support the separately announced VMware Private AI Foundation.

NVIDIA AI Ready Servers are equipped with NVIDIA® L40S GPUs, NVIDIA BlueField®-3 DPUs, and NVIDIA AI Enterprise Software. This enables enterprises to customize and deploy generative AI applications, such as intelligent chatbots, search, and summarization tools, using proprietary business data.

NVIDIA L40S-based servers are set to be released at the end of the year through Dell Technologies' PowerEdge R760xa, Hewlett Packard Enterprise's ProLiant Gen11, Lenovo's ThinkSystem SR675 V3, and others.

The L40S GPU handles complex AI workloads with billions of parameters. It includes 4th generation Tensor Cores and an FP8 Transformer Engine, providing over 1.45 petaflops (PF) of tensor processing performance and up to 1.7 times the training performance compared to the NVIDIA A100 Tensor Core GPU.

Equipped with NVIDIA BlueField DPUs, it accelerates virtualization, networking, storage, security, and other cloud-based AI services. NVIDIA ConnectX®-7 SmartNICs provide best-in-class scalability for data-intensive generative AI workloads.

NVIDIA AI Enterprise Software provides production support and security for NVIDIA's more than 100 frameworks, pre-trained models, toolkits, and software for LLM.
NVIDIA AI Ready Servers provide full-stack accelerated infrastructure and software, enabling their use in various applications such as new drug development, product description, intelligent virtual assistants, and manufacturing simulation.

Jensen Huang, founder and CEO of NVIDIA, stated, "A new era of computing has begun. Companies across all industries are competing to adopt generative AI. NVIDIA is providing generative AI to companies worldwide through its ecosystem of the world's leading software and systems partners."

Michael Dell, CEO of Dell Technologies, said, “Generative AI is a catalyst for innovation and helps solve the world’s most pressing challenges.” "We are pleased to be able to accelerate the deployment of generative AI by partnering with NVIDIA to integrate GPUs and software into various enterprise tuning and inference workload solutions," said Antonio Neri, CEO of HPE.

Meanwhile, NVIDIA AI Ready servers can build a VMware Private AI Foundation. "By combining the expertise of VMware, NVIDIA, and our server manufacturer partners, enterprises can develop and deploy AI while maintaining data privacy, security, and control," said Raghu Raghuram, CEO of VMware.

NVIDIA AI Ready Servers are scheduled for release at the end of the year. They will be available through cloud service providers within the next few months.