Nvidia’s RTX Spark laptops are now available, and early testing results are promising. Most reviews suggest strong GPU performance, decent software compatibility, and laptops that feel like premium, MacBook-style devices.
But as Extreme Tech reports, they are very expensive, and outside AI developers, most PC users won’t need such powerful local AI machines.
Nvidia promised earlier this year that the new RTX Spark laptops would run everything well. That’s largely proven accurate, with The Verge reporting that most apps do indeed run well. It certainly runs games with anti-cheat software like Fortnite, Rust, and Roblox, despite those proving problematic for SteamOS and other Windows-on-Arm systems.
CPU-limited games perform quite badly, though, because while the N1 and N1X chips offer strong multi-threaded performance, their single-threaded capabilities aren’t great, according to HotHardware. Combined with the need to feed CPU instructions through a Windows translation layer first, CPU-bound games will be limited to 30 frames per second, even at higher graphical settings.
Tom’s Hardware describes the local AI experience as fast and effective, with the N1X laptop outperforming a MacBook M5 and Ryzen AI Halo on the same complex workloads. However, it says gaming is far from impressive, citing the shared memory bandwidth of just 300GB/s as particularly shoddy—less than you’d get on an RTX 5060 mobile, for example.
That’s why the laptop redlines its chips to output only 1080p, and stutters and hitches aren’t uncommon.
Windows Central is more forgiving, noting this is the first time we’ve seen RTX 5060-like performance from an SoC, but it agreed it’s optimized for AI workloads and lacks capabilities elsewhere. The CPU is “just OK,” though the GPU really shines, it said.
Over at Tom’s Guide, a reviewer found gaming performance in some games particularly troubling, with 1% lows dipping well below 20fps. A laptop with this kind of price tag shouldn’t struggle. But AI-wise, the N1X is clearly capable, with lots of promise for future efficiency gains and optimizations.
—
Photo Credit: DANIEL CONSTANTE / Shutterstock.com