Nvidia Announced 3 New Powerful AI Chips Last Week

Nvidia's conference was inspirational for the future of Artificial Intelligence hardware

Nvidia’s Annual Conference wouldn’t usually be noteworthy, the but implications in A.I. hardware meant, I just had to cover it.

I think it’s even safe to say that Nvidia is primed to capture the exponentially growing AI market with its three key advantages:

• Omniverse: A digital twin of the real world that obeys the laws of physics and allows simulation of real world challenges in a virtual environment before building the physical.

• CUDA: A programming language platform allowing developers to build their own AI apps.

• GPUs for data centers: The powerful chips which enable high performance computing needs for AI and deep learning.

During his keynote address, Nvidia CEO Jensen Huang did say that Eos (Nvidia’s new supercomputer), when running traditional supercomputer tasks, would rack 275 petaFLOPS of compute — 1.4 times faster than “the fastest science computer in the US” (the Summit).

If you haven’t seen it yet, I highly recommend you check out the intro to the GTC 2022 keynote with Jensen Huang, I found it pretty moving. If you don’t see it below you can watch it here.

It has over 7 million views. (view below)

It’s getting harder to track all that is occuring with A.I. chips in industry.

Nvidia’s A.I. Hardware Dominance

It’s a bit daunting to cover the hardware advances of Nvidia, since they are like custodians of what the data-center and GPU has become.

Nvidia Corp on Tuesday announced several new chips and technologies that it said will boost the computing speed of increasingly complicated artificial intelligence algorithms, stepping up competition against rival chipmakers vying for lucrative data center business.

Nvidia’s Strange Blend of A.I. Evangelism and Visionary Innovation is Enchanting

Nvidia is building a world of robo-taxis and bigger A.I. language models that will power the future of tomorrow’s society.

Nvidia announced its new H100 AI processor, a successor to the A100 that's used today for training artificial intelligence systems to do things like translate human speech, recognize what's in photos and plot a self-driving car's route through traffic.

AI is powering change in every industry across the globe. As companies are increasingly data-driven, the demand for AI technology grows. From speech recognition and recommender systems to medical imaging and improved supply chain management, AI technology is providing enterprises the compute power, tools, and algorithms their teams need to do their life’s work.

• Nvidia’s Omniverse and work with digital twin technology is astounding to me.

NVIDIA Omniverse™ is a scalable, multi-GPU real-time reference development platform for 3D simulation and design collaboration, and based on Pixar's Universal Scene Description and NVIDIA RTX™ technology.

It might become an unparalleled B2B place for industry collaboration among developers to build the next-gen of digital transformation and digital worlds. I believe it may have significant impacts on the industrial metaverse and whatever it becomes.

I find Nvidia’s “I am AI” series moreover quite persuasive on A.I’s ability to do good in the world. It’s the best corporate A.I. for good campaign I know of. In comparison, I find Microsoft’s public relations and Google’s rhetoric on A.I. really laughable.

Do you enjoy A.I. articles at the intersection of breaking news, then help me continue to write on the subject. I cannot continue to write without support. Grateful for all tips, patronage and community contributions.

NVIDIA Introduces Grace CPU Superchip

• 144 High-Performance Cores and 1 Terabyte/Second Memory; Doubles Performance and Energy-Efficiency of Server Chips.

Nvidia’s GPUs are dominant in A.I. Workloads in 2022

Nvidia's graphic chips (GPU), which initially helped propel and enhance the quality of videos in the gaming market, have become the dominant chips for companies to use for AI workloads.

• The latest GPU, called the H100, can help reduce computing times from weeks to days for some work involving training AI models, the company said.

“A new type of data center has emerged — AI factories that process and refine mountains of data to produce intelligence,” said Jensen Huang, founder and CEO of NVIDIA.

I think you could say A.I. hardware is keeping up with increasingly large A.I. language models and even pushing them forwards.

On A.I. Evangelism and Its Corporate Use

Nvidia is doubling down on artificial-intelligence technology that CEO Jensen Huang predicts will revolutionize every industry. It’s just more believable when Jensen is saying it, and not Sundar, Satya, Mark or whoever else. Jensen actually sounds like a real person.

A.I. is Everywhere Narrative is Occuring

Companies have been using AI and machine learning for everything from making recommendations of the next video to watch to new drug discovery, and the technology is increasingly becoming an important tool for business.

• The A100 is based on a design called Hopper, and it can be paired with Nvidia's new central processing unit, called Grace, Nvidia said at its GPU Technology Conference.

• Nvidia has benefited hugely from the AI boom of the last decade, with its GPUs proving a perfect match for popular, data-intensive deep learning methods. As the AI sector’s demand for data compute grows, says Nvidia, it wants to provide more firepower.

• In particular, the company stressed the popularity of a type of machine learning system known as a Transformer. This method has been incredibly fruitful, powering everything from language models like OpenAI’s GPT-3 to medical systems like DeepMind’s AlphaFold.

• Such models have increased exponentially in size over the space of a few years. When OpenAI launched GPT-2 in 2019, for example, it contained 1.5 billion parameters (or connections). When Google trained a similar model just two years later, it used 1.6 trillion parameters.

• Over the past few years, a number of companies with strong interest in AI have built or announced their own in-house “AI supercomputers” for internal research, including Microsoft, Tesla, and Meta.

The H100 GPU itself contains 80 billion transistors and is the first GPU to support PCle Gen5 and utilize HBM3, enabling memory bandwidth of 3TB/s.

Nvidia Chips are Well Diversified in our Lives

Nvidia has a broad collection of processor designs touching many parts of our digital lives. Nvidia may not be as well known as Intel or Apple when it comes to chips, but the Silicon Valley company is just as important in making next-gen technology practical.

Nvidia also announced the RTX A5500, a new member of its Ampere series of graphics chips for professionals who need graphics power for 3D tasks like animation, product design and visual data processing.

Importance of TMSC: Taiwan Manufacturing Semiconductor Company

The H100 chip will be produced on Taiwan Manufacturing Semiconductor Company's cutting edge four nanometer process with 80 billion transistors and will be available in the third quarter, Nvidia said.

Tl;dr

Nvidia has announced a slew of AI-focused enterprise products at its annual GTC conference. They include details of its new silicon architecture, Hopper; the first datacenter GPU built using that architecture, the H100; a new Grace CPU “superchip”; and vague plans to build what the company claims will be the world’s fastest AI supercomputer, named Eos.

Nvidia’s EOS Supercomputer

The H100 will also be used to build Nvidia's new "Eos" supercomputer, which Nvidia said will be the world's fastest AI system when it begins operation later this year.

How to conceptualize the scale of Eos? I’m not a hardware guy.

The supercomputer, named Eos, will be built using the Hopper architecture and contain some 4,600 H100 GPUs to offer 18.4 exaflops of “AI performance.” The system will be used for Nvidia’s internal research only, and the company said it would be online in a few months’ time.

• Eos is based on the fourth-generation DGX system — the DGX H100 — that was also launched today, and which is powered by octuple NVLink-connected H100 GPUs (more on all of that here).

• An external NVLink switch can then connect these DGX H100s into Pods, which offer up to an exaflop of AI performance and can themselves be linked in 32-node increments to form systems like Eos.

In total, Eos (see rendering picture above) will contain 18 of these 32-DGX H100 Pods, for a total of 576 DGX H100 systems; 4,608 H100 GPUs; 500 Quantum-2 InfiniBand switches; and 360 NVLink switches. “Eos will offer an incredible 18 exaflops of AI performance, and we expect it to be the world’s fastest AI supercomputer when it’s deployed,” said Paresh Kharya, senior director of product management and marketing at Nvidia, in a prebriefing for press and analysts.

Conclusion

Nvidia unveiled three of these systems, each pushing the limits of computing power and designed to work together to help unlock the potential of the AI economy.

Nvidia’s vision of the future of the data center really does feel futuristic.

Huang said Hopper is not just an update to Ampere (although new Nvidia technology is always compatible with older generation hardware). Hopper solves new problems and will achieve breakthroughs in what AI and machine learning are capable of doing.

A.I. hardware is powering the future of machine learning in a quickly evolving world.

Nvidia's ability to help its customers process massive amounts of data to power AI is well known now in 2022.

Grace CPUs are still expected to be available during the first half of 2023. 

I hope that gives you a quick summary of how impactful Nvidia’s conference was this Spring and how hardware is powering the future in the space.

If you thought this article was interesting and could be valuable to someone, feel free to share it. I try to keep the majority of my articles free due to my paid contributors (who show me tips, patronage and community support).

Do you enjoy A.I. articles at the intersection of breaking news, then help me continue to write on the subject. I cannot continue to write without support. Grateful for all tips, patronage and community contributions.

So by subscribing you are essentially helping fund a network of Newsletters whose aim is to inspire and inform. This is my only job and stream of income.

AiSupremacy is the fastest Substack Newsletter in AI at the intersection of breaking news. It’s ranked #1 in Machine Learning as of January 22nd, 2022.

Thanks for reading!