Nvidia's New Chip: Smartphone Moment for PCs?
Nvidia just dropped the RTX Spark at Computex 2026, and CEO Jensen Huang isn't being subtle about the stakes, he's comparing it to the invention of the smartphone.
The chip promises to turn your Windows laptop into an AI "teammate," not just a tool, with major OEMs like Dell, HP, Lenovo, and Microsoft Surface already on board for autumn releases.
But here's where it gets interesting: Forrester's Charlie Dai calls this a "paradigm shift" from Nvidia being a component supplier to becoming an "architecture owner" in the PC market.
That's a direct shot at Intel, AMD, Qualcomm, and Apple. Meanwhile, analysts are already flagging a "significant price tag," suggesting this won't be for everyday consumers anytime soon.
And then there’s the elephant in the room - the same week Nvidia unveils the consumer push, the US tightened export rules on Nvidia's advanced Blackwell chips to Chinese firms. A global chip war is quietly escalating while Jensen Huang takes selfies on stage.
Is this genuinely the next big computing revolution, or just expensive AI hype dressed up in thin aluminium?
Personally, this feels like Nvidia wants developers hooked on its ecosystem, not just its hardware.
Replies
Uh oh - NVIDIA beating AMD or Intel to the punch (let's hope they respond). I think NVIDIA is too powerful - I wonder what the DOJ or Federal Trade Commission thinks about this. First I head about this but I'll look into it now.
As a dev, I don't care about the 'teammate' marketing fluff. I care about whether I can run a 70B model locally without melting my power supply. If this gets developers to stop relying on the cloud for inference, that's a genuine shift. But let’s be honest—Nvidia knows that once a dev writes their code for Nvidia’s specific architecture, they aren't switching to AMD or Intel next generation. The price is just the entry fee for that walled garden. I’m more worried about the export restrictions; it means the AI race is now officially two separate tracks, and we’re about to see two different computing standards emerge.