Michael M. Santiago
Nvidia (NASDAQ:NVDA) kicked off its annual GTC conference with a bevy of announcements from Chief Executive Jensen Huang’s keynote address that left many on Wall Street impressed.
Shares slipped 1.5% in mid-day trading on Tuesday.
“It is clear from the GTC keynote that despite all that has happened in the last 18 months, Nvidia believes we are still early when it comes to unlocking the potential of generative AI,” Morgan Stanley analyst Joseph Moore wrote in a note. Moore has an Overweight rating and $795 price target on Nvidia.
While the majority of the keynote was focused on developers, the company laid out a grand ambition of an “AI-powered future,” with generative artificial intelligence “the building block of a new industrial revolution,” Moore said.
“It will take time to evaluate the performance claims for Blackwell, but if they hold up even directionally, our sense is that the company’s ability to raise the bar this much leaves them in a very strong position,” Moore added. “While most of the larger cloud customers remain committed to alternative AI solutions – custom or merchant – they all have limited rack space given ecosystem constraints (with multiple hyperscalers waiting to power up new facilities). Limited rack space is going to lead to cloud vendors choosing the highest ROI solution, which we continue to think is represented by NVIDIA.”
AMD (AMD), which competes with Nvidia in the AI accelerator market, fell 6% on Tuesday, while Intel (INTC), Qualcomm (QCOM), Broadcom (AVGO) and others were down modestly.
Announcements aplenty
Among the many announcements made by Huang include the company’s new Blackwell GPU platform, named after American mathematician David Blackwell. The new platform contains the B100 and B200 GPUs, which are twice the size of the H100.
Amazon (AMZN) Web Services, Google (GOOG) (GOOGL) Cloud, Microsoft (MSFT) Azure and Oracle (ORCL) Cloud Infrastructure will be among the first cloud service providers to offer Blackwell-powered instances, Nvidia said. Applied Digital, CoreWeave, Crusoe, IBM (IBM) Cloud and Lambda will also offer Blackwell support.
Other announcements include the GB200, new X800 series networking switches (designed for massive-scale AI), Nvidia’s next AI supercomputer, the NVIDIA DGX SuperPOD and new NVLink chip.
Although most of the announcements were seen as in-line with expectations, KeyBanc Capital Markets analyst John Vinh said the event — and the updates to the Blackwell architecture — were “nothing short of amazing.”
Nvidia expects the average selling price for the B100 to be roughly 40% higher than the H100, which starts at $35,000, while its AI performance is between 2.5 and 5 times greater than the H100.
Huang told CNBC that the B200 will cost between $30,000 and $40,000.
All of these announcements have reinforced Nvidia’s leadership over companies such as AMD (AMD), Intel (INTC) and others in the AI space and it’s likely that the increase in average selling price and performance boost will help sustain “outsized earnings growth,” Vinh added.
Other analysts also praised Nvidia after the keynote address, including Wedbush analyst Matt Bryson, which reiterated his Outperform rating and $1,000 price target. Wells Fargo analyst Aaron Rakers boosted his price target to $970 from $840 after the event, noting the increased focus on Omniverse (which now supports Apple’s (AAPL) Vision Pro) and digital twins may be just as important.
“While the Nvidia Omniverse 3D rendering / simulation platform may be materializing a bit slower than initially expected, WELLS continues to view as an underappreciated piece of Nvidia’s long-term AI-platform strategy / monetization opportunity,” Rakers wrote in a note to clients.
(This story has been updated to include pricing details on the B200.)
