8 Recent Nvidia Updates: Nutanix AI Deal, AI Data Center Guidelines And More

The AI computing giant’s recent moves include expanded partnership with hybrid multi-cloud infrastructure vendor Nutanix, new reference architectures for AI data centers and new peer-reviewed performance numbers for its upcoming Blackwell GPUs.

Nvidia may have not announced a new GPU over the past three months, but that doesn’t mean it wasn’t busy finding other ways to expand and cement its high influence in the fast-growing AI computing market.

For his part, Nvidia CEO and co-founder Jensen Huang has been busy trotting the globe and meeting with world leaders with industry titans recently to tout new AI supercomputer deals underlining the company’s belief in what become known as “sovereign AI.”

This refers to the idea that nations must build out their own AI infrastructure, workforces and business networks to advance their economies, according to Nvidia.

“Denmark recognizes that to innovate in AI, the most impactful technology of our time, it must foster a domestic AI infrastructure and ecosystem,” Huang said in an October summit with King Frederik X of Denmark regarding a new AI supercomputer built in the country.

But while Huang has been making stops in Denmark and other countries like Japan, India and Indonesia, his company has been making moves in other ways in recent months.

These moves include an expanded partnership with hybrid multi-cloud infrastructure vendor Nutanix, new reference architectures for AI data centers, new peer-reviewed performance numbers for its upcoming Blackwell GPUs, the hiring of a prolific Cisco Systems investor and Blackwell-related contributions to the Open Compute Project.

What follows are summaries of these and other recent updates from Nvidia, organized in reverse-chronological order from when they were announced.

Nvidia: Blackwell GPU Double LLM Performance In Peer-Reviewed Tests

Nvidia said its forthcoming B200 Blackwell GPU is two times faster than its previous-generation H100 GPU in large language model fine-tuning and pre-training.

This is according to new peer-reviewed results in the latest MLPerf Training suite of benchmark tests released Wednesday by MLCommons.

The AI computing giant said the B200 GPU is 2.2 times faster for fine-tuning with the 70-billion-parameter Llama 2 model and two times faster for pre-training with the 175-billion-parameter GPT-3 model compared to the H100 on a per-GPU basis.

Nvidia said the large language model (LLM) tests were performed using an AI supercomputer called Nyx that is built with the company’s DGX B200 systems, each of which come with eight 180-GB B200 GPUs.

The company also highlighted how ongoing software updates result in performance and feature improvements for its H100 GPU, which debuted in 2022 and enabled the “highest performance amongst available solutions” in the new MLPerf Training tests.

Among the results submitted in the latest MLPerf Training test suite, Nvidia said it demonstrated that the H100 achieved a 30 percent boost in training for the GPT 175B model compared to when the benchmark was initially released.

It also demonstrated the impact of software improvements in other tests.

“Nvidia Hopper GPUs have more than tripled scale and performance on the GPT-3 175B benchmark since last year. In addition, on the Llama 2 70B LoRA fine-tuning benchmark, Nvidia increased performance by 26 [percent] using the same number of Hopper GPUs, reflecting continued software enhancements,” the company said.

Nutanix Expands Partnership With Nvidia For New Cloud AI Offering

Nutanix announced on Tuesday that it has expanded its partnership with Nvidia through a new cloud-native offering it’s calling Nutanix Enterprise AI.

The hybrid multi-cloud infrastructure vendor said Nutanix Enterprise AI can be “deployed on any Kubernetes platform, at the edge, in core data centers, and on public cloud services” such as Amazon Web Services’ Elastic Kubernetes Service, Microsoft Azure’s Managed Kubernetes Service, and Google Cloud’s Google Kubernetes Engine.

“With Nutanix Enterprise AI, we're helping our customers simply and securely run GenAI applications on-premises or in public clouds. Nutanix Enterprise AI can run on any Kubernetes platform and allows their AI applications to run in their secure location, with a predictable cost model,” said Thomas Cornely, senior vice president of product management at Nutanix, in a statement.

Nutanix said the cloud-native offering can be deployed with Nvidia’s full-stack AI platform and is validated to work with the Nvidia AI Enterprise software platform, including Nvidia NIM microservices that are designed to enable “secure reliable deployment of high-performance AI model inferencing.”

Nutanix Enterprise AI is part of the company’s GPT-in-a-Box 2.0 appliance, which it said is part of the Nvidia-Certified System program to ensure optimal performance.

Nvidia Replaces Intel In Dow Index

Nvidia replaced Intel on the Dow Jones Industrial Average on Nov. 8 as the AI computing giant continues to put competitive pressure on the beleaguered chipmaker.

S&P Dow Jones Indices, the organization behind the DJIA, said in a statement the week before that the change is meant “to ensure a more representative exposure to the semiconductors industry.”

While Nvidia’s stock price is up more than 200 percent from the beginning of the year, Intel’s shares have been down roughly 40 percent across the same period.

“The DJIA is a price weighted index, and thus persistently lower priced stocks have a minimal impact on the index,” according to S&P.

Nvidia Adds Veteran Astronaut Ellen Ochoa To Board

Nvidia announced near the beginning of November that it has appointed veteran astronaut and former NASA director Ellen Ochoa to its board of directors.

The AI computing giant said the addition of Ochoa, who was the first Latina astronaut in space and the former director of NASA’s Johnson Space Center in Houston, expanded the board to 13 members.

“Ellen’s extraordinary experience speaks volumes about her role as a trailblazer and a leader,” Nvidia CEO Jensen Huang said in a statement. “We look forward to her joining Nvidia’s board on our continuing journey to build the future of computing and AI.”

Nvidia Reveals Enterprise Reference Architectures For AI Data Centers

Nvidia on Oct. 29 revealed what it’s calling Enterprise Reference Architectures, which are meant to serve as guidelines for partners and customers building AI data centers.

The AI computing giant said these guidelines are meant to help partners and customers build AI data centers faster and achieve business value from AI applications sooner while ensuring they operate at optimal performance levels in a secure manner.

Server vendors participating in the Enterprise Reference Architectures program include Dell Technologies, Hewlett Packard Enterprise, Lenovo and Supermicro.

The Enterprise Reference Architectures consist of a comprehensive set of recommendations for accelerated infrastructure, AI-optimized networking and the Nvidia AI Enterprise software platform.

The accelerated infrastructure guidelines provide recommendations for GPU-accelerated servers in the Nvidia-Certified Systems program, which ensure that such servers are optimized to deliver the best performance.

The networking guidelines address how to “deliver peak network performance” with Nvidia’s Spectrum-X Ethernet platform and BlueField-3 data processing units. They also offer “guidance on optimal network configurations at multiple design points to address varying workload scale requirements.”

Nvidia AI Enterprise includes NeMo and NIM microservices for “easily building and deploying AI applications” as well as Nvidia Base Command Manager Essentials for “infrastructure provisioning, workload management and resource monitoring.”

Nvidia Contributes Blackwell Platform Design To Open Compute Project

Nvidia on Oct. 15 announced that it has contributed “foundational elements” of its Blackwell accelerated computing platform to the Open Compute Project with the goal of enabling wide adoption across the data center market.

The Blackwell platform in question is Nvidia’s upcoming GB200 NVL72 rack-scale system that comes with 36 GB200 Grace Blackwell Superchips, each of which contain an Arm-based, 72-core Grace GPU connected with two B200 GPUs.

Nvidia’s contributions to the Open Compute Project (OCP) focus on the GB200 NVL72’s electro-mechanical design. The elements contributed by Nvidia include the rack architecture, the compute and switch tray mechanicals, the liquid-cooling and thermal environment specifications as well as the NVLink cable cartridge volumetrics.

The AI computing giant said it has also expanded support for OCP standards of its Spectrum-X Ethernet platform. This will let customers use Spectrum-X’s “adaptive routing and telemetry-based congestion control to accelerate Ethernet performance for scale-out AI infrastructure,” according to the company.

“Building on a decade of collaboration with OCP, NVIDIA is working alongside industry leaders to shape specifications and designs that can be widely adopted across the entire data center,” Nvidia CEO Jensen Huang (pictured) said in a statement. “By advancing open standards, we’re helping organizations worldwide take advantage of the full potential of accelerated computing and create the AI factories of the future.”

Nvidia Hires Top Cisco Inventor Amid Big Networking Sales Push

Nvidia recently hired a 25-year Cisco Systems engineering veteran, once credited as the switching giant’s most prolific inventor, who will lead development of AI and networking architecture at the AI computing giant.

JP Vasseur, a recently departed Cisco fellow who was most recently vice president of engineering of machine learning and AI for networking, announced in a LinkedIn post on Oct. 2 that he joined Nvidia last month as a senior distinguished engineer and chief architect of AI and networking.

Vasseur’s announcement comes a little more than a month after Nvidia CFO Colette Kress said that the company’s Spectrum-X line of Ethernet networking products for data centers is “well on track to begin a multibillion-dollar product line within a year.”

Nvidia Launches NIM Agent Blueprints To Boost Enterprise AI

Nvidia on Aug. 27 announced the launch of “pre-trained, customizable AI workflows” that can help enterprises more easily build and deploy custom generative AI applications.

The AI computing giant is calling these workflows NIM Agent Blueprints, which come with sample applications built with Nvidia NeMo, Nvidia NIM and partner microservices as well as reference code, customization documentation and a Helm chart for deployment.

The channel partners chosen by Nvidia to offer NIM Agent Blueprints consist of Accenture, Deloitte, SoftServe and World Wide Technology.

Nvidia revealed the microservices at its GTC 2024 event in March as an addition to its Nvidia AI Enterprise software platform. At the time, the company said they are meant to help businesses develop and deploy AI applications faster “while retaining full ownership and control of their intellectual property.”

With NIM Agent Blueprints, enterprises can customize generative AI applications using their own proprietary data and continuously refine them based on user feedback.

The initial set of NIM Agent Blueprints include workflows for digital humans in customer service, generative virtual screening in computer-aided drug discovery and multimodal PDF extraction in retrieval-augmented generation.