PHOTO: Nvidia.
NVIDIA unveiled their latest range of chips designed to power artificial intelligence yesterday at a developer conference in California, according to AFP. This move comes as they aim to solidify their position as the leading supplier in the booming AI market.
“We need bigger GPUs,” declared CEO Jensen Huang, addressing the audience. “So, ladies and gentlemen, I’d like to introduce you to a very, very big GPU,” he added, referencing the graphics processors crucial for developing generative AI.
The event, nicknamed “AI Woodstock” by Wedbush analyst Dan Ives, has become a key fixture in the big tech calendar. This is thanks to Nvidia’s unique role in the AI revolution that has swept the world since the introduction of ChatGPT in late 2022.
“Just a reminder, folks,” Huang quipped as he took the stage in the bustling arena, more accustomed to ice hockey and gigs than tech talks. “This ain’t a concert, it’s a developers’ conference.”
Nvidia’s Dominance in AI Hardware
Nvidia’s brawny GPU chips and software are the linchpin for building generative AI, AFP reports. Competitors like AMD and Intel are still grappling to match the power and efficiency of Nvidia’s smash-hit H100 product, which launched in 2022.
Apple, Microsoft, and Amazon have all dabbled in developing AI-focused chips, but for now, they’re stuck trying to get their hands on Nvidia’s sought-after tech to fulfil their own AI ambitions.
That lynchpin role in the AI revolution has seen Nvidia’s share price rocket roughly 250% over the past year, propelling them above Amazon in terms of market cap, with only Microsoft and Apple ahead.
Nvidia Unveils “Blackwell” Superchips
Not one to rest on their laurels, Nvidia unveiled an even more powerful processor and accompanying software to a captivated audience of developers and tech execs, according to AFP. The platform is called Blackwell, named after David Blackwell, the first Black academic inducted into the National Academy of Sciences.
These Blackwell GPUs are billed as AI “Superchips,” boasting four times the speed of the previous generation when training AI models, Nvidia claims.
“The pace of advancement in computing is mind-blowing,” declared Huang.
Crucially, they’ll also deliver 25 times the energy efficiency, Nvidia claims. This is a key point considering the criticism AI development faces for its guzzling of energy and resources compared to traditional computing.
Unlike rivals like Intel, Micron and Texas Instruments, Nvidia, along with AMD, adopts a fabless model – they don’t manufacture their chips. Instead, they rely on subcontractors, primarily the Taiwan Semiconductor Manufacturing Company (TSMC).
Given the current geopolitical tensions between Taiwan and China, this could be a potential vulnerability. It’s worth noting that the US has already banned Nvidia from exporting its most powerful chips to Chinese companies.
Further AI Developments from Nvidia
According to AFP, Nvidia also announced other AI developments, including a platform for training humanoid robots. Project Gr00t, which Nvidia claims has no connection to the Guardians of the Galaxy character of the same name, was described as the “world’s first human foundation model.” Robots powered by Gr00t will be designed to understand human speech and mimic human movements, learning how to interact with the world through experience, Nvidia explained.
These models “will enable a robot to learn from a small number of human demonstrations, allowing it to assist with everyday tasks and emulate human movement simply by observing us,” Nvidia said.
AFP also reported that Nvidia is collaborating with Apple to integrate AI capabilities into the newly released Vision Pro spatial computing gear. This partnership comes as Apple faces pressure to demonstrate it’s keeping pace with Amazon, Google, Meta, and OpenAI in the field of artificial intelligence.
Finally, Nvidia unveiled the Earth-2 Cloud Platform, designed to predict climate change using simulations run by AI supercomputers.
-BTS Media
