More

    Nvidia Reveals GeForce RTX 50 Series GPUs, Launching January 2025

    Coming to desktops, shelves, and laptops in the next few months.

    During CES 2025, Nvidia CEO Jensen Huang revealed the company’s lineup of new GeForce RTX 50 series GPUs, launching on January 30 with the RTX 5090 and 5080. These new GPUs will support DLSS 4.0 Multi Frame Generation, which is a major upgrade over the DLSS 3.0 series of machine learning super sampling technology.

    Unlike previous generation RTX GPUs, GeForce RTX 50 series GPUs are “SFF-Ready” with two-slot models. This means that they’re thin enough to fit inside ITX builds. Every GPU features GDDR7 video memory, DLSS 4.0 support, 5th generation tensor cores, 4th generation ray tracing cores, and support for 4K at 480Hz / 8K at 120Hz. The 16-pin power connector is still in use as well.

    DLSS 4.0 uses roughly 30% less video memory, and 75 games are expected to support it thus far, including Alan Wake 2, Cyberpunk 2077, Indiana Jones and the Great Circle, Star Wars Outlaws, DOOM: The Dark Ages, and more. Nvidia also showed off the new Virtua Fighter project, but it was in-engine footage and not actual gameplay.

    GeForce RTX 50 Series GPU Specifications (To Be Updated)

    Nvidia GeForce RTX 5090

    The flagship model includes:

    • 21,760 CUDA Cores
    • 680 5th Generation Tensor Cores
    • 170 4th Generation Ray Tracing Cores
    • 32 GB of GDDR7 Video Memory
    • 512-bit Memory Bus
    • 1792 GB/sec Total Memory Bandwidth
    • 2.01 GHz Base Clock
    • 2.41 GHz Boost Clock
    • 575W TDP
    • 1000W Recommended System Power
    • Price: $1,999
    Nvidia GeForce RTX 5080

    The 80 class GPU includes:

    • 10,752 CUDA Cores
    • 336 5th Generation Tensor Cores
    • 84 4th Generation Ray Tracing Cores
    • 16 GB of GDDR7 Video Memory
    • 256-bit Memory Bus
    • 960 GB/sec Total Memory Bandwidth
    • 2.30 GHz Base Clock
    • 2.62 GHz Boost Clock
    • 360W TDP
    • 850W Recommended System Power
    • Price: $999
    Nvidia GeForce RTX 5070 Ti
    • 8,960 CUDA Cores
    • 280 5th Generation Tensor Cores
    • 70 4th Generation Ray Tracing Cores
    • 16 GB of GDDR7 Video Memory
    • 256-bit Memory Bus
    • 896 GB/sec Total Memory Bandwidth
    • 2.30 GHz Base Clock
    • 2.45 GHz Boost Clock
    • 300W TDP
    • 750W Recommended System Power
    • Price: $749
    Nvidia GeForce RTX 5070
    • 6,144 CUDA Cores
    • 192 5th Generation Tensor Cores
    • 48 4th Generation Ray Tracing Cores
    • 12 GB of GDDR7 Video Memory
    • 192-bit Memory Bus
    • 672 GB/sec Total Memory Bandwidth
    • 2.16 GHz Base Clock
    • 2.51 GHz Boost Clock
    • 250W TDP
    • 650W Recommended System Power
    • Price: $549

    The flagship GPU, the RTX 5090, will be available on January 30 alongside the RTX 5080. The RTX 5070 Ti and RTX 5070 GPUs will both launch this February. Nvidia will release Founders Edition models of the RTX 5090, RTX 5080, and RTX 5070 graphics cards worldwide.

    Stock and factory-overclocked variants of all four GPUs will be available from top board partners such as ASUS, Colorful, Gainward, GALAX, GIGABYTE, INNO3D, KFA2, MSI, Palit, PNY and ZOTAC. System builders like Northwest, Maingear, Origin PC, and more will feature the GPU. 

    GeForce RTX 50 series GPUs will also be available in laptops starting this March. More details about DLSS 4.0 and GeForce RTX 50 series laptops may be read here and here at Nvidia’s official website.

    Soul Kiwami
    Soul Kiwami
    Raul Ochoa, a.k.a. Soul Kiwami, is the Managing Editor of Final Weapon and a Games Writer at Game Rant with four years of writing and editing experience. Raul is passionate about the Japanese gaming industry, and he's a huge fan of Nintendo Switch, PC hardware, JRPGs, and fighting games. business email: [email protected] | Muck Rack: https://muckrack.com/raul-landaverde-1

    Latest articles

    Latest Articles