Topics
Latest
AI
Amazon
Image Credits:Justin Sullivan / Getty Images
Apps
Biotech & Health
Climate
Image Credits:Justin Sullivan / Getty Images
Cloud Computing
Commerce
Crypto
Image Credits:Justin Sullivan / Getty Images
Enterprise
EVs
Fintech
Image Credits:David Paul Morris / Bloomberg / Getty Images
fund raise
contrivance
game
Government & Policy
ironware
layoff
Media & Entertainment
Meta
Microsoft
Privacy
Robotics
security system
Social
Space
startup
TikTok
Transportation
speculation
More from TechCrunch
outcome
Startup Battlefield
StrictlyVC
newssheet
Podcasts
Videos
Partner Content
TechCrunch Brand Studio
Crunchboard
get hold of Us
Nvidia take San Jose by tempest this year , with a record - breaking 25,000 meeter flocking to the San Jose Convention Center and surrounding downtown buildings . Many shop , talks , and panels were so jam-packed that mass had to lean against walls or sit on the floor — and suffer the wrath of organizers shouting commands to get them to run along up properly .
Nvidia presently sit at the top of the AI world , withrecord - breaking financials , sky - mellow profit margin , and no serious contender yet . But the coming months also check unprecedented danger for the company as it confront U.S. tariffs , DeepSeek , and pitch priority from top AI customers .
AtGTC 2025 , Nvidia CEO Jensen Huang attempted to project confidence , unveilingpowerful new scrap , personal “ supercomputers,”and , of course of study , really cute robot . It was an exhaustive sales pitching — one aimed at investor reeling from Nvidia’snosediving pedigree .
“ The more you buy , the more you bring through , ” Huang said at one point during a tonic on Tuesday . “ It ’s even better than that . Now , the more you buy , the more you make . ”
Inference boom
More than anything , Nvidia at this year ’s GTC assay to secure attendees — and the rest of the world watching — that need for its chips wo n’t slacken down anytime before long .
During his tonic , Huangclaimed that nearly the “ entire world get it wrong”ontraditional AI scale return out of vogue . Formosan AI laboratory DeepSeek , which sooner this yearreleased a highly efficient “ reasoning ” modelcalledR1 , prompted veneration among investors that Nvidia ’s monster chips may no longer be necessary for training militant AI .
But Huang has repeatedly insist that power - hungry reasoning model will , in fact , tug more requirement for the ship’s company ’s splintering , not less . That ’s why at GTC , Huang showed offNvidia ’s next line of Vera Rubin GPUs , claim they ’ll perform illation ( that is , run AI models ) at roughly double the charge per unit of Nvidia ’s current salutary Blackwell chip .
Join us at TechCrunch Sessions: AI
Exhibit at TechCrunch Sessions: AI
The threat to Nvidia ’s business that Huang spent less time come up to included upstarts like Cerebras , Groq , and other small - cost inference hardware and swarm provider . Nearly every hyperscaler is developing a custom chip for illation , if not training , as well . AWS has Graviton and Inferentia ( which it ’s reportedly aggressively discounting ) , Google has TPUs , and Microsoft hasCobalt 100 .
Along the same venous blood vessel , tech giants presently exceedingly reliant on Nvidia chips , including OpenAI and Meta , are looking to reduce those ties viain - househardwareefforts . If they — and the said other rivals — are successful , it ’ll almost assuredly weaken Nvidia ’s throttlehold on the AI chip market .
That ’s perhaps whyNvidia ’s share price dippedaround 4 % following Huang ’s keynote . Investors might ’ve been contain out Leslie Townes Hope for “ one last affair ” — or perhaps an accelerated launch windowpane . In the end , they got neither .
Tariff tensions
Nvidia also essay to ease worries about duty at GTC 2025 .
The U.S. has n’t imposed any tariffs on Taiwan ( where Nvidia gets most of its chip ) , and Huangclaimedtariffs would n’t do “ significant legal injury ” in the short campaign . He discontinue short of assure that Nvidia would be shielded from the retentive - term economic impacts , however — whatever physical body they in the end take .
Nvidia has clearly have the Trump administration ’s “ America First ” message , with Huangpledging at GTCto spend hundreds of billion of dollars on manufacturing in the U.S. While that would help oneself the companionship diversify its supplying chains , it ’s also a massive price for Nvidia , whose multitrillion - dollar valuation calculate on healthy profit margin .
New business
As it looks to source and maturate businesses other than its gist chip melodic line , Nvidia at GTC drew attention to its new investments in quantum , an industry that the company has historically miss . At GTC ’s first Quantum Day , Huang apologizedto the CEOs of major quantum companies for causing a minor lineage clangour in January 2025 after he suggested that the technical school would n’t be very useful for the next 15 to 30 year .
On Tuesday , Nvidiaannouncedthat it would open a fresh center in Boston , NVAQC , to make headway quantum computing in coaction with “ leading ” hardware and software package markers . The sum will , of trend , be equipped with Nvidia chips , which the company says will enable researchers to simulate quantum system of rules and the good example necessary for quantum error chastening .
In the more immediate future tense , Nvidia sees what it’scalling “ personal AI supercomputers”as a possible young revenue - maker .
At GTC , the companylaunchedDGX Spark ( antecedently calledProject Digits ) and DGX Station , both of which are designed to allow users to prototype , fine - tune , and run AI fashion model in a range of sizes at the bound . Neither is exactly cheap — they retail for thousands of dollar mark — but Huang boldly proclaimed that they represent the hereafter of the personal PC .
“ This is the data processor of the age of AI , ” Huang said during his keynote . “ This is what computers should look like , and this is what computers will run in the future . ”
We ’ll shortly see if customers check .
