NVIDIA Launches Four RTX 50 Series GPU: RTX 5070 Debuts at $549, RTX 5070 Ti at $749, RTX 5080 at $999, and Flagship RTX 5090 at $1999
NVIDIA’s CEO Jensen Huang held the keynote at CES, unveiling the latest Blackwell GeForce RTX 50 series GPUs. NVIDIA revealed four GPUs in the lineup, unveiling their specifications, performance, and exclusive features that bring a new leap in gaming performance through the Blackwell architecture.
The RTX 50 series brings forward the flagship GeForce RTX 5090 GPU, bringing 2X the performance of the RTX 4090 with DLSS 4.0. The RTX 5090 is the first gaming GPU from NVIDIA to bring 32GB of GDDR7 memory on a 512-bit bus, delivering 1792 GB/s of memory bandwidth. Equipped with 21,760 Cuda Cores, the GPU drastically enhances its capabilities over predecessors.
The RTX 5090 is packed with the 5th generation 680 Tensor Cores, and 4th generation 170 Ray Tracing Cores to deliver significantly better raster and ray tracing performance. In games like Cyberpunk 2077, Alan Wake 2, and Black Myth: Wukong, NVIDIA RTX 5090 with DLSS 4.0 will be ideal for playing at 4K 240Hz monitors for an immersive and detailed gaming experience.
Apart from gaming, the RTX 5090 delivers excellent performance in generative AI applications, delivering the same results in half the time compared to the RTX 4090. As revealed in the presentation, the RTX 5090 Founders Edition card is a dual-fan design with just a 2-slot thickness, measuring 304mm x 137mm for better compatibility with PC towers. The RTX 5090 is officially priced at $1999 and is the most expensive flagship 90-class card till now.
Apart from the GeForce RTX 5090, NVIDIA also unveiled the GeForce RTX 5080 as the second fastest GPU in the lineup, bringing two times the performance of RTX 4080 with DLSS 4.0. The GPU brings 5th-gen Tensor and 4th-gen RT Cores, 16GB GDDR7 VRAM on a 256-bit bus, delivering a good 960 GB/s of memory bandwidth. The GeForce RTX 5080 is priced at $999, which was the launch MSRP of the RTX 4080 Super.
The third in the lineup is the GeForce RTX 5070 Ti, featuring 16GB GDDR7 memory on a 256-bit bus, delivering up to 896 GB/s of memory bandwidth. The GPU is roughly 2X faster than the RTX 4070 Ti with DLSS 4.0 and is priced at $749.
The fourth is the GeForce RTX 5070, bringing 12GB GDDR7 memory on a 192-bit memory bus, resulting in about 672 GB/s of memory bandwidth. This is about 33% higher memory bandwidth than the RTX 4070. NVIDIA claims that the RTX 5070 is perfect for 1440p resolution and with full ray tracing and DLSS 4.0 multi-frame generation enabled, it can deliver very high frame rates. The GeForce RTX 5070 is priced at $549.
All the announced GPUs will be available in Q1 2025, starting with RTX 5090 and RTX 5080, which will be launched for retail stores on 30th January. The RTX 5070 Ti and RTX 5070 will be available in February. Apart from the desktop RTX 50-series GPUs, NVIDIA also launched the Laptop RTX 50 series GPUs as well. These include the GeForce RTX 5090 Laptop GPU, featuring 10,496 Cuda Cores, 24GB GDDR7 VRAM, and 1824 AI TOPs for high computing performance.
The GeForce RTX 5080 Laptop GPU features 7,680 Cuda Cores, 16GB GDDR7 VRAM, and 1334 AI TOPs, resulting in about 2X the performance of the RTX 4080 laptop GPU. The RTX 5070 laptop GPU family includes the GeForce RTX 5070 Ti Laptop GPU and GeForce RTX 5070 Laptop GPU. The RTX 5070 Ti Laptop GPU features 5888 Cuda Cores, 12GB GDDR7 VRAM, and 992 AI TOPs while the RTX 5070 Laptop GPU offers 4608 Cuda Cores, 8GB GDDR7 VRAM, and 798 AI TOPs.
Laptops with NVIDIA GeForce RTX 50-series mobile GPUs will be available starting in March except RTX 5070-based laptops, which will launch in April.
