Nvidia chases $200B CPU market with AI agent PCs from Microsoft, Dell, and HP

2 days ago 8

Nvidia opened Taipei’s tremendous Computex commercialized amusement connected Sunday with a spark, literally. The chipmaker unveiled a caller PC CPU called the RTX Spark, which it dubbed a “superchip,” and named a who’s who database of PC makers that volition soon present AI PCs powered by it.

The super-fast, 1-petaflop spot is designed to tally AI agents similar OpenClaw oregon Hermes Agent securely, according to Nvidia. Such RTX Spark Windows PCs volition beryllium disposable this autumn from ASUS, Dell, HP, Lenovo, Microsoft Surface and MSI, with models from Acer and Gigabyte to follow.

In summation to being equipped with unafraid sandboxes (jointly developed with Microsoft) to tally agents securely, the PCs volition besides person capable CPU, GPU, RAM and underlying Nvidia CUDA bundle to tally section versions of ample connection models.

Nvidia said that its RTX exertion volition present faster show for AI, amended representation quality, and enactment for AI features successful much than 1,000 games and applications.

The chipmaker is selling this arsenic an alternate for creators making AI content, arsenic good arsenic providing a important upgrade to its accepted marketplace of gamers. Nvidia said much than 100 Windows bundle makers person signed connected to enactment the caller chip, including Adobe, Blender, ComfyUI, Riot Games and Xbox.

But Nvidia laminitis and CEO Jensen Huang’s imaginativeness for these caller PCs is acold larger. He wants to extremity the days of launching apps, pointing, clicking and typing.

“With RTX Spark and Microsoft Windows, you inquire — and the PC does the work,” helium said successful the property release. “Frontier models. Creative workflows. RTX games. All connected a laptop.”

Last month, aft delivering different grounds quarter, Huang promised investors helium had recovered a caller $200 cardinal marketplace for Nvidia successful selling CPUs for AI, not conscionable GPUs. He made circumstantial notation of the high-end server CPU released earlier this twelvemonth called Vera — of which Nvidia says it has already sold $20 cardinal worth.

He besides hinted astatine his bigger ambitions. “We’ll person billions of agents, and those billions of agents volition each usage tools. And those tools are going to beryllium similar PCs, conscionable similar america humans utilizing using PCs today,” helium said connected the net telephone successful May. “We’re going to request a batch much CPUs.”

Nvidia ARM-based Windows devices person been tried earlier — and failed. Back successful 2013, Microsoft famously had to constitute disconnected $900 million connected its Nvidia ARM-based Surface RT, with partners similar Dell besides bailing connected the product.

But astatine this point, aft delivering grounds aft grounds of quarterly revenue, it’s hard to stake against Huang arsenic helium pursues his PC dreams erstwhile again.

And this spot is an wholly antithetic beast. It’s much powerful, not less. Microsoft is positioning its ain RTX Spark PC arsenic truthful mighty that it named it the Surface Laptop Ultra, and is calling it “the astir almighty Surface Laptop ever built.”

Still, PC manufacturers person not released a batch of specifics astir each of their offerings, including pricing. These systems look to beryllium full-fledged Windows versions of the DGX Spark mini-computer that Nvidia already sells to developers for astir $4,800.

We’ll person to hold and spot if these PCs volition vie connected terms with the affordable Mac Mini that has go a fashionable prime for moving OpenClaw. Or possibly they volition beryllium astatine the precocious extremity of the PC market, similar Nvidia’s ain agent-running mini computer.

Either way, if Nvidia has cracked the codification connected bringing AI agents easily, safely, and usefully to the masses, it could — and should — beryllium big.

When you acquisition done links successful our articles, we whitethorn gain a tiny commission. This doesn’t impact our editorial independence.

Read Entire Article