wasl.tech

Nvidia RTX Spark AI PCs Land in India: Laptops From Rs 2.99 Lakh, 128GB Memory

What Nvidia Showed at the RTX Spark Experience

Picture a video editor in Mumbai, cutting 8K footage on a laptop in a café, no charger in sight. Or a developer in Bengaluru, running a 74-billion-parameter AI model on the machine in front of her, with no cloud bill ticking in the background.

Until recently, both scenes needed either a heavy workstation or a fast internet connection. Nvidia now says neither is required.

At its RTX Spark Experience event in India on Friday, Nvidia showed off a new class of AI-focused Windows PCs built around its RTX Spark superchip. Five brands — Asus, Dell, HP, Lenovo and MSI — brought slim laptops and compact desktops to the stage, all designed to run serious AI work locally, on the device itself.

According to Gadgets 360, the systems combine a Blackwell RTX graphics processor with a Grace central processor, offering up to 128GB of unified memory and up to one petaflop of FP4 AI performance. It is the same silicon Nvidia has been previewing since the start of the year, now arriving in machines Indian buyers can actually order.

RTX Spark AI PCs: The Lineup and India Prices

The cheapest way into the new lineup comes from Lenovo. Its Yoga Pro 9n starts at Rs. 2,99,399, while the Yoga 9n 2-in-1 convertible starts at Rs. 3,32,801. Both are slim laptops aimed at creators and professionals who want local AI without carrying a desktop replacement.

Asus brought two ProArt machines. The ProArt P14 starts at Rs. 3,19,990 and the larger ProArt P16 at Rs. 4,89,990. The company’s compact ProArt GR1X desktop, the same mini PC that went on sale globally earlier in the week, starts at Rs. 9,00,000 in India.

Dell’s entry is the XPS 16 Creator Edition, priced from Rs. 4,49,990. It went on sale through Dell’s online store on November 20, the same day as the event. The laptop is just 17.8mm thin and carries a 3.2K tandem OLED display, a screen technology Dell usually reserves for its flagship machines.

HP’s OmniBook Ultra starts at Rs. 5,49,999, with configurations offering up to 128GB of memory. Pre-bookings open on November 24 through HP’s online store. MSI’s EdgeMesa N AI+ compact desktop starts at Rs. 3,89,999, making it the most affordable desktop in the showcase.

One more machine is still on the way. Gigabyte’s AI TOP Aero is expected to launch in India later in December, starting at Rs. 7,35,499. Nvidia also confirmed that Acer and HP mini PCs are coming, along with Microsoft’s Surface RTX Spark Dev Box for developers.

One Superchip, 128GB of Shared Memory

Every machine in the lineup is built around the same core idea: one superchip instead of a separate processor and graphics card. The RTX Spark pairs a 20-core Grace CPU, designed with MediaTek, with a Blackwell RTX GPU carrying 6,144 CUDA cores. The two share a single pool of up to 128GB of unified memory.

That shared memory is the point. In a conventional laptop, system RAM and graphics memory are separate, and a large AI model has to be squeezed into whichever pool is smaller. Here, the whole 128GB is available to both, so demanding creative applications and local AI models can run at the same time without fighting over memory.

The result, Nvidia says, is up to one petaflop of FP4 AI performance in a machine that fits in a messenger bag. The company also claims all-day battery life on the laptops, and says the systems can handle 8K 4:2:2 video editing and render 3D scenes of up to 90GB, even when running on battery power.

It is worth remembering that these are Nvidia’s own claims. No independent benchmarks of the shipping systems have been published yet, and the global mini PC launch earlier this week arrived with the same caveat. Real-world tests will tell whether the battery and performance promises hold up.

Local AI Agents Without the Command Line

The most interesting part of the showcase was not the hardware at all. It was what the machines are meant to do when they are not rendering video or playing games: run AI agents locally, on the device, without sending your data to a data centre.

Nvidia says users can start with agents such as GitHub Copilot, Hermes, OpenClaw, Perplexity, Pi and WorkBuddy through guided setup tools and installers, without touching a command line. The pitch is that the PC stops being a tool you operate and becomes a teammate that completes tasks, generates assets and writes code.

The company also highlighted Nemotron 3.5 Bolt 74B-A9B 24GB, a 74-billion-parameter model optimised for tool calling and multi-step tasks, for supported RTX Spark configurations. A model that size would normally need a server rack; here it is meant to run on a laptop.

There is a practical side to this, too. A machine that serves AI inference to other computers on a home network is part of the plan, through a feature Nvidia calls PAIR.

It lets an RTX Spark system handle AI work for compatible Windows, Linux and macOS computers nearby, while keeping those other machines free for creating and gaming.

Security and the Software Stack

Running agents on your own machine raises an obvious question: what stops an agent from doing something you did not ask for? Nvidia’s answer is a secure runtime called OpenShell, combined with Microsoft’s Execution Containers, which give users control over agent permissions.

Compatible agents can also interact with tools such as Nvidia Broadcast, RTX Remix and Omniverse libraries. And the full Nvidia software stack comes along for the ride: CUDA, DLSS, OptiX, Reflex, G-SYNC, TensorRT and hardware-accelerated video, all in the same slim chassis.

That software depth is the real advantage over the first wave of so-called AI laptops. Most of those machines bolted a small neural processor onto a conventional design and called it a day. RTX Spark is a different bet: a data-centre-style architecture, shrunk down and sold as a personal computer.

What It Means for Buyers in India

The prices put these machines firmly in premium territory. At Rs. 2,99,399, the cheapest laptop costs roughly what a good mid-range car costs to insure for a year, and the top configurations stretch past Rs. 9,00,000. These are not machines for everyone.

But the audience is clear. Creators editing 8K video, developers fine-tuning models, researchers running simulations and studios rendering 3D scenes are the people Nvidia is chasing. For them, the promise is simple: workstation-class AI work, anywhere, without a cloud subscription.

The timing matters as well. The RTX Spark desktops from Asus, Dell and MSI went on sale globally on November 18, and the India event three days later shows how quickly Nvidia wants the platform in front of buyers. With Acer, Gigabyte, HP and Microsoft machines still to come, the lineup will only grow from here.

Nvidia’s chip ambitions reach well beyond laptops. The company recently committed $1 billion to Anthropic and OpenAI, a reminder that the same AI boom driving these PCs is reshaping the whole industry.

For now, the first machines are on sale or about to be. Dell’s XPS 16 is available today, HP’s pre-bookings open on November 24, and Gigabyte’s Aero follows in December. The question for buyers is whether local AI is worth premium money — and the first independent reviews will answer that better than any launch event can.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top