Nvidia GeForce RTX 5070 Laptop vs Nvidia GeForce RTX 5060 Laptop vs Nvidia GeForce RTX 5080 Laptop

Nvidia GeForce RTX 5070 Laptop

The Nvidia GeForce RTX 5070 Laptop is a mobile graphics card from the Blackwell family for laptops, which was unveiled at CES 2025. It is based on the GB206 chip with 4,608 shader / CUDA cores clocked at 2235 - 2520 MHz (depending on the model). The graphics memory used is 8 GB GDDR7 with a 128-bit memory bus.

The chip is manufactured in the current TSMC 4N process.

A new feature of the Blackwell architecture is support for multi-frame generation (MFG) in DLSS4.

Note: All of the above information is based in part on leaks such as TechPowerUp.

Nvidia GeForce RTX 5060 Laptop

The Nvidia GeForce RTX 5060 Laptop is an upcoming Blackwell family graphics card for laptops that Team Green is very likely to unveil during CES 2025 with retail availability expected by April 2025.

Considering the 4060 Laptop has 3,072 unified shaders (CUDA cores) at its disposal, the 5060 will most most likely sport over 4,000 shaders making most 2025 and 2024 games playable at 1080p on Ultra. Nvidia might tap one of Samsung's or TSMC's 3 nm processes to make the chip but 4 nm processes are not off the table, either.

Other features of the GPU include native PCIe 5 support, DisplayPort 2.1a, 8 GB of 128-bit GDDR6 or GDDR7 VRAM and a 115 W max power target (not including Dynamic Boost - ?).

Note: All of the above is based on various leaks and speculations like this one here (VideoCardz).

Nvidia GeForce RTX 5080 Laptop

The Nvidia GeForce RTX 5080 Laptop is a high-end notebook graphics card that was presented at CES 2025 and will go on sale in April 2025. It uses the GB203 Blackwell chip and offers 8,192 of the chip's 10,752 cores. The graphics memory used is 16 GB GDDR7, which is connected via a 256-bit memory bus (896 GB/s maximum bandwidth).

A new feature of the Blackwell architecture is support for multi-frame generation (MFG) in DLSS4.

The GB203 is manufactured at TSMC using the 4N FinFET process (4nm or 5nm). Currently, much of the data is still based on the specifications at Techpowerup.