We’re live at Nvidia‘s Computex 2026 keynote, where CEO Jensen Huang will share his vision of the future for not just the company, but the computing industry as a whole.
While the company used to be best known for creating some of the best GPUs ever, it’s also one of the leading names in AI, and that has meant it’s become one of the most valuable companies in the world.
This has meant that in recent keynotes, Huang has spent an increasing amount of time talking about artificial intelligence, and less time on graphics cards.
This year is shaping up to be the same, as the AI hype shows little sign of slowing down, and there haven’t been any rumors or leaks to suggest Nvidia is about to pull a new GPU out of its hat.
However, there have been rumors circulating that Nvidia could launch a new Arm-based chip for laptops at Computex 2026, directly competing with the likes of Apple and Qualcomm (who both make chips based on Arm architecture), as well as CPU heavyweights Intel and AMD.
With Nvidia, Microsoft, and Arm all tweeting hints that this year’s Computex is one to watch, it’s looking increasingly likely that Nvidia could announce its N1X chip, which could seriously shake up the computing landscape.
For the first time in years, the Nvidia keynote is one you don’t want to miss, and we’re in the audience at the Taipei Music Center in Taiwan to bring you all the breaking news as it happens.
You can also watch the live stream of the keynote below:
NVIDIA GTC Taipei 2026 Keynote | Live – YouTube
- Check out our Computex 2026 hub for all the latest news out of Taipei.
Refresh
Good morning from Taipei, Taiwan! I’ve taken a seat ready for Nvidia’s keynote presentation, and the hall is filling up quickly. We’re about an hour away from Jensen Huang taking to the stage.
More leaks have emerged overnight about Nvidia’s rumored Arm-based chip for laptops, so we could be in for a treat during this keynote. Hopefully there won’t be too much AI and data center stuff…
As you can see, the hall in Taipei Music Center, where the keynote will take place, is already nearly full, 45 minutes before the presentation starts.
Here I am with my press pass. Apologies for the dishevelled look, I’m a bit jet lagged. It’s currently 10:16am in Taiwan.
One thing I’m not expecting at this year’s keynote is a new GPU. There hasn’t been much talk of a successor to Nvidia’s 50 series graphics cards, and considering the current memory shortage, which has seen price rises for existing graphics cards, I don’t think there’s much appetite from Nvidia to launch a new GPU any time soon.
As we reported last month, there’s also a feeling that Nvidia is losing interest in graphics cards. While the company is clearly making much more money from AI, it would be a real shame if it stopped making GPUs.
I can’t see that happening myself, but if this keynote barely mentions graphics cards, and instead focuses on AI and the rumored N1X chip, then those fears won’t disappear.
I wonder if we’ll hear and see more about DLSS 5? Its announcement was certainly controversial, with critics saying that it’s little more than an AI beauty filter that detracts from artists’ original intent in games. There was quite the backlash, so it’ll be interesting to see if Huang addresses this, and possibly demonstrates the technology to give us a better idea of what it does, or if he doesn’t bring it up at all…
We are about 10 minutes out from the keynote starting…
If you want to watch along at home, Nvidia is live streaming the keynote on YouTube. I’ll also embed the video at the top of this page.
OK, one minute to go! I’m also going to try to record some of this. While typing this live blog. Wish me luck! And apologies for any typos!
It begins with a moody video about… AI tokens. Yey.
Jensen Huang is onstage! He has a cute shoutout to his parent who are in the audience.
He’s thanking Taiwanese partners (and getting laughs).
‘Useful AI has arrived’ according to Huang. Hmm, I’m not too convinced…
Talking about GitHub commits and how they are growing rapidly. I guess the point is that this growth is powered by AI.
A lot of code commits, sure. But are they any good?
‘People talk about AI cutting jobs – complete nonsense’ says Huang. He suggests AI is increasing coding jobs, not reducing them.
AI is now a profit generator, according to Huang. As a journalist, I’m not sure that applies to all industries…
Huang is now talking about AI agents and how a simple prompt can turn into code, which can then create an application.
‘This is a great time to be a software company’ he says. Again, claims AI will create jobs, not get rid of them. I’m not convinced, and the muted audience suggests I’m not alone.
Now a video showing off computer animation and gaming effects, such as RTX Mega Geometry. Video has a disclaimer saying none of this is generative AI.
‘Math is beautiful’ says Huang after the video wraps up. Nerd!
He’s talking about full-stack AI infrastructure, and to be honest I’ve lost a bit of interest. I want to see some consumer hardware!
‘A long time ago, Nvidia used to be a GPU company. We evolved.’
Now a video about AI factories. I have nothing to say about this.
Sorry, if that sounded a bit grumpy, despite the jet lag I’m not that grumpy! But this video is about tech that I am sure is very impressive, but it’s not my area of expertise so can’t really comment on it.
All these AI factories that Nvidia is promising sounds good – but where will the power come from? AI and data centers are already putting a strain on energy infrastructure in many countries.
The video suggests these AI factories will be efficient, but all this talk of gigawatts doesn’t seem to suggest that these energy concerns will go away.
Huang is now pitching to companies directly, it feels. ‘Compute is revenue’ he says, and warns companies not to pick the wrong platform (IE not Nvidia) just because ‘the chips are cheaper’.
A video about Vera Rubin AI platform, which is now in production. Huang thanks the audience for making this happen. Uh, you’re welcome I guess. Not sure how I helped with this…
Huang said ‘Vera Rubin’ so many times I got a bit annoyed. Oops, grumpy Matt has popped up again!
‘This is an agentic system’. Huang is talking about how important agents are. He’s bringing out Vera Rubin now.
Here it is.
‘I think there are 2,000 people behind there, pulling it,’ quips Huang as the Vera Rubin is wheeled off stage…
I think I’m going to hear ‘Vera Rubin’ in my head over and over again after this. He keeps saying it.
He’s talking about SQL running three times faster on the Vera CPU.
Huang’s got out his laser pointer. It’s green. We are still on AI infrastructure.
Now he’s talking about reinventing PC with Microsoft…
Introducing RTX Spark!
128GB RAM.
Reinventing the personal computer for creators and gamers.
This is the Arm chip that was rumored,
The Nvidia RTX Spark comes with 20 CPU cores (the CPU has been custom-designed by Nvidia and MediaTek) and 6,144 CUDA cores based on Blackwell architecture.
Lots of laptops coming in ‘Fall’.
He claims this will be as revolutionary as the change from phones to smartphones. Big claim. All down to agentic AI.
Says ‘100%’ of PC world is behind this. Not sure that includes Apple, Intel, AMD, Qualcomm and other chip making rivals…
He’s now talking about AI in cars, so that’s the chip announcement done. RTX Spark laptops will come later this year, but no mention of price. I think these will be very expensive. Huang will be appearing at Microsoft Build tomorrow. Good news is that we’ll also be at that event!
Keynote is winding down now with ‘one more thing’, which is a video recap.
Oh god it’s a cringy video with a singing robot and what looks like generative AI. 🙁
There’s a song about what was announced. It’s. Uh, not great.
I’ll wrap up this live blog now, as I need to dash to more Computex events, keep an eye on TechRadar for all our Computex 2026 coverage, and tune in to Build 2026 tomorrow for more information about RTX Spark!
