Ill-Advised Shunt Mod Can Give GeForce RTX Laptops A Massive FPS Boost
The gains? Substantial—but only if you do the work. In 3DMark's Steel Nomad benchmark, the stock 5090 config scored 6404 points. At 225W, it went to 6895 (+7.7%), and at 250W, 7098 (+10.8%). With manual overclocking (including undervolting and power curve tuning) on top of that, it hit a monstrous 9404 points, a 41% uplift from stock and faster than a desktop GeForce RTX 5080. Shunt modding unlocks the wattage, but real gains still demand overclocking know-how.
While GizmoSlipTech argues that NVIDIA should raise laptop GPU power limits to compete with Apple's efficient silicon, that feels like a stretch. You can certainly improve performance by throwing more power at the problem; that's exactly what NVIDIA has done the last two generation with its "x90" desktop GPUs. That's not really a great solution in a laptop form factor, though; in the end, you're creating something more like one of the old "luggable" PCs rather than a proper laptop.
What's actually worth noticing is how much performance NVIDIA is leaving on the table. Even with top-end hardware, the RTX 5090 mobile GPU is clearly bottlenecked by its 175W cap. Our own testing of the ASUS ROG Strix G16 with a far smaller RTX 5070 Ti shows it coming surprisingly close to RTX 5090 laptops in some real-world benchmarks. That's a red flag, or at least a sign that NVIDIA's current power policies are keeping high-end silicon on arguably too tight a leash.
In the end, the shunt mod story isn't about whether you should do it, because you almost assuredly shouldn't. It does shine a light on something NVIDIA might not want you to know, though: its most powerful laptop GPUs have a lot more headroom than the company is letting on. While we're not as enthusiastic as GizmoSlipTech about it—mostly because existing desktop replacements are already massive—we definitely agree that it would be cool for manufacturers to have more freedom in designing ultra-high-powered gaming portables.
