In his first live keynote speech since the pandemic, company co-founder and chief executive officer (CEO) Jensen Huang spoke for nearly two hours about the “accelerated computing services, software and systems” that he said are creating new business models and making current ones more efficient.
“We are now at the tipping point of a new computing era, with accelerated computing and AI that’s been embraced by almost every computing and cloud company in the world,” he said.
A key launch revolved around the Nvidia MGX server specification, which the company described as a means for system manufacturers to cost-effectively build more than 100 server variations to suit a range of artificial intelligence (AI), high-performance computing (HPC), and Omniverse applications. The latter is Nvidia’s 3D graphics collaboration platform for building and operating metaverse applications.
According to a company release, with MGX, manufacturers start with a basic system architecture optimized for accelerated computing as their server chassis, and then “select their GPU, DPU and CPU. Design variations can address unique workloads, such as HPC, data science, large language models, edge computing, graphics and video, enterprise AI, and design and simulation.
“Multiple tasks like AI training and 5G can be handled on a single machine, while upgrades to future hardware generations can be frictionless. MGX can also be easily integrated into cloud and enterprise data centres.”
Nvidia announced that SoftBank Corp. plans to use MGX to dynamically allocate GPU resources between generative AI and 5G applications as it rolls out multiple hyperscale data centres across Japan.
“As generative AI permeates across business and consumer lifestyles, building the right infrastructure for the right cost is one of network operators’ greatest challenges,” said SoftBank president and CEO Junichi Miyakawa. “We expect that Nvidia MGX can tackle such challenges and allow for multi-use AI, 5G and more, depending on real-time workload requirements.”
John Annand, a director of the infrastructure team at Info-Tech Research, said that Huang made a good point in his keynote, namely that “AI is not just about the chips. The software supporting the silicon is as, or even more, important. Nvidia has a long history of partnerships with highly technical vendors.
“HPC, self-driving vehicles, AR/VR, rendering, engineering simulation, analytics, computer vision, you name a deep technical field, and chances are Nvidia has a product evangelist hoeing that row and working side by side with the pioneers in that specialty.”
This, he said, gives them a definite edge in the sense that “whatever fabrication or engineering advances Intel, AMD, or Broadcom can make to try and leapfrog Nvidia GPUs, they still have to develop the presence, awareness, and software for the market to embrace them.”
Of note is that both company revenues and share price are soaring. Nvidia shares were trading at US$398.50 this afternoon, and earlier this week the company’s market value reached the coveted US$1 trillion mark, before declining slightly, a massive achievement.
The big angle, said Annand, is what impact Nvidia’s current success is going to have on Intel and AMD: “My first big takeaway is that a rising tide lifts all boats. Nothing breeds copycats like success. Rumors are that AMD and Microsoft are collaborating on Project Athena, which is supposedly a new line of AI-based chips.
“Broadcom is a major silicon manufacturer announcing their own “general purpose” AI chips, working with vendors like Google, to say nothing of their own ‘specific-AI’ chips like the Jericho3, which helps connect up to 32,000 GPUs together.”
Annand added that, at Innovation 2022, Intel, “to much fanfare, had a computer vision demonstration with Chipotle as a partner. Edge devices deployed to real live restaurant locations that monitor freshness and stock levels of produce and food on a line to reduce waste. Nvidia’s success will undoubtedly cause other silicon manufacturers to double down on their fast-follow efforts.”
Nvidia also launched the following this week at Computex:
- Nvidia Spectrum X, an accelerated networking platform the company said is designed to improve the performance and efficiency of Ethernet-based AI clouds. “Transformative technologies such as generative AI are forcing every enterprise to push the boundaries of data center performance in pursuit of competitive advantage,” said Gilad Shainer, senior vice president of networking at Nvidia. “Nvidia Spectrum-X is a new class of Ethernet networking that removes barriers for next-generation AI workloads that have the potential to transform entire industries.”
- The DGX GH2000 AI supercomputer, powered by the firm’s GH200 Grace Hopper Superchips and the Nvidia NVLink Switch System, which was created, the company said, “to enable the development of giant, next-generation models for generative AI language applications, recommender systems and data analytics workloads.”