LG claims the lightest Nvidia RTX laptop to date

LG is banking on featherlight power. After getting hands-on with the latest additions to the Gram family at its show-floor booth, the headline is obvious: The Gram Pro 17 with Nvidia GeForce RTX graphics feels incredibly portable, and LG says it’s the lightest RTX laptop yet. The wider lineup gravitates around a new alloy and AI-first features, capturing the sense that the Grams were a shot across performance’s bow… without suffering backache.

Why LG’s ‘lightest RTX laptop’ claim stands out

The Gram Pro 17 packs a 17-inch 1600p display with a discrete Nvidia GeForce RTX 5050 GPU into a body that’s only 0.63 inches thick and weighs 3.84 pounds. I placed a pre-production unit on one of those compact travel scales and landed right around that number. (For context, most RTX-powered 17-inch systems here today still tend to clump around 5 to 6 pounds, and even many of the hassle-free ultraportable performance models in this class end up closer to a shade above 4 pounds.) Dipping under 4 but achieving a large screen and RTX silicon isn’t common.

LG pitches the claim as “the lightest RTX notebook” in general, but it’s meaningful in use: creators and shoulder warriors can now toss a true dGPU system into a messenger bag without worrying about a trip to a chiropractor on Monday. Until we’re able to get retail units in-store, and rivals respond, it’s just a claim—but on initial hands-on evidence it makes a good first impression.

The new lightweight alloy that makes these cuts possible

All the models move over to Aerominum, LG’s fancy magnesium-aluminum alloy. The material feels upscale—matte, cool to the touch—and is helping with weight cuts all around. LG claims that the chassis has been “tested to meet U.S. military standards of durability” (MIL-STD-810H is a benchmark used by many PC makers; LG doesn’t specify here which tests it took). I could flex it by picking up the corner of the 16-inch model, but day-to-day sturdiness seemed solid for something so thin.

Four finishes are offered—white, black, silver, and a particularly snappy bronze that seems to say “executive” without tipping into bling. The fit and finish appeared to be the same from unit to unit, which is good news for mass production.

Gram Pro 16 hits the ultralight portability sweet spot

The Gram Pro 16 is the crowd-pleaser. It measures a mere 0.64 inches thick and, critically, only 2.64 pounds—well under the unwritten 3-pound barrier that has marked any laptop bestowed with the moniker “ultraportable” for years. That’s saying something for a 16-inch laptop that features a full-size keyboard and numpad. The 1600p OLED-looking panel was punchy and accurate; we saw inky blacks and crisp text visible to my naked eye under bright hall lighting.

Inside, look for options from Intel’s next-gen Core Ultra family (code-named Panther Lake) to AMD’s Ryzen AI chips. I was told by LG that the AMD models will have a separate “Gram Book” branding to differentiate them from the Intel SKUs. And despite the thinness, the port setup is actually practical: USB-C, HDMI, and a headphone jack all return. There’s a 2-in-1 version coming for the traditional pen input and convertible feel for people.

I’ll be keeping an eye on how well these review models can maintain performance and thermals—ultralights tend to excel for a bit and then flatten out under longer loads.

That said, for those who commute, switch from conference rooms to meetings, or run through code or content workflows that spike and rest, the Pro 16’s balance looks compelling.

Gram Pro 17 brings RTX power without the usual bulk

Re-View: The Gram Pro 17 is the poster child of LG’s RTX assertion. The GeForce RTX 5050, on the other hand, is aimed at creators who depend on GPU acceleration for duties such as timeline scrubbing, AI upscaling, and 3D previews. You can also game more casually on the 1600p panel if you tweak settings. For those who don’t need the extra silicon, LG will have lighter 17-inch versions sans discrete graphics.

Physics is always the clincher in ultra-thin designs, so cooling and acoustics will be the two variables to watch. Anticipate conservative power targets versus thicker workstations; that’s frequently just fine, anyway, for mobile-first creators who value silent operation and battery-friendly efficiency over peak frame rates.

AI-first PCs with fast local features and smart cloud help

All the new Grams are tagged “AI” and are therefore Copilot+ PCs, having a dedicated NPU onboard to accelerate on-device experiences. On a day-to-day level, that potentially means quicker, private tasks such as removing backgrounds from images, transcribing text, and tweaking photos without ever waking the cloud. Microsoft has been championing that paradigm, and hardware support is now rapidly moving from being a value option to being a baseline in premium laptops.

LG brings its own software contributions with Dual AI. There is the local assistant Gram Chat, which can operate offline to summarize documents, navigate settings, and look for files in a secure folder. Beyond that, when you need even more horsepower or better reasoning, the LG Cloud AI draws on models as strong as OpenAI’s GPT-4o and GPT-4o mini. That hybrid approach reflects where the industry is going—run what you can locally for speed and privacy, then burst to the cloud when quality is what matters most.

Early verdict: who these new LG Gram laptops are for

If you’ve been biding your time for a big-screen laptop that won’t take its toll on your shoulders, this Gram generation is perhaps the most convincing yet. The Pro 16 seems like a default recommendation for travelers and students in the market for a spacious OLED that doesn’t exceed 3 pounds, and the Pro 17 puts RTX acceleration across a category that once started at closer to five pounds.

Exactly how long the battery holds up, how loud the fans get when they wind up for gaming, and how well the RTX 5050 maintains performance in such a thin shell are open questions, but the materials jump and AI-first stance add up to a clear strategy. LG isn’t pursuing the fastest benchmark numbers; it’s pursuing effortless portability with just enough headroom for modern creation and AI workloads. That sounds pretty spot-on from what I saw and considered.