To achieve RTX 4090 Laptop GPU performance NVIDIA GeForce RTX 4080 Laptop GPU requires 20W more power. ComputerBase evaluates the power scaling of the newest mobile graphics from NVIDIA. The concern of lower-tier SKUs outperforming flagship SKUs as they have more power headroom is well-established. This became a significant concern two GeForce generations ago, requiring NVIDIA and laptop manufacturers to reveal more details on the TGP (Total Graphics Power) of each RTX Laptop GPU installed in each system.
However, buyers who do not consider the TGP statistics may end up overpaying for power-limited high-end GPU cards. In this situation, one might have just as easily purchased a less expensive laptop with a more powerful GPU. ComputerBase evaluated two laptops, the MSI TITAN GT77HX V13 and the XMG Neo 16 E23, both of which were integrated with RTX 4090/i9-13950HX and RTX 4080/i7-13700HX graphics cards. The maximum TGP configuration for the GPU in each system is 150W plus 25W for Dynamic Boost (power dynamically allocated between CPU and GPU). Each GPU’s power scaling was tested on these platforms.
To illustrate the performance per watt scaling in a single graph, they used Shadow Warrior 3 with 1440p resolution and maximum details. The RTX 4090’s minimum TGP power setting is 80W, while the RTX 4080’s minimum TGP power setting is 60W. There is a possibility that RTX 4080 will be faster than RTX 4090; nevertheless, it will need 20W more power to achieve RTX 4090 efficiency. Although the flagship GPU from NVIDIA has improved power efficiency, this may not always compensate for the higher cost of the laptop.
The power scaling of the AD104, AD106, and AD107 GPUs, which will be released later this month, should be significantly more intriguing. Most high-end laptops currently feature RTX 4080/4090 with high TGP variations, so the likelihood of purchasing a significantly throttled system, particularly with RTX 4090 graphics, is low.