Nvidia's GeForce RTX 5090 is equipped with 92 billion transistors and it is capable of delivering 3,352 trillion operations per second (TOPS).
Nvidia's new GeForce RTX 50 series GPUs will be available later this month
Photo Credit: Nvidia
Nvidia on Monday announced the launch of its GeForce RTX 50 Series GPUs at CES 2025. The company's high-end graphics cards are equipped with up to 32GB of GDDR7 memory and 21,760 CUDA cores, and feature Nvidia's latest deep learning super sampling (DLSS 4) technology. The top-of-the-line RTX 5090 GPU is capable of delivering 3,352 trillion operations per second (TOPS), while the company claims that the successor to last year's RTX 4090 model outperforms the high-end GPU by "up to 2x".
The flagship GeForce RTX 5090 GPU is priced at Rs. 2,14,000, while the RTX 5080 model costs Rs. 1,07,000. Meanwhile, the GeForce RTX 5070 Ti and RTX 5070 will set you back by Rs. 80,000 and Rs. 59,000, respectively.
The company says the RTX 5090 and RTX 5080 models will be available on Jan 30, while the RTX 5070 Ti and RTX 5070 GPUs will be available starting in February. They will be available from brands such as Asus, Colorful, Gainward, Galax, Gigabyte, INNO3D, KFA2, MSI, Palit, PNY, and Zotac.
Nvidia GeForce RTX 5090
Photo Credit: Nvidia
Customers can also expect laptops equipped with the GeForce RTX 5070 GPU to arrive by April, while models featuring the other three models will be available starting in March. The GPUs will be part of laptops from Acer, Asus, Dell, Gigabyte, HP, Lenovo, MECHREVO, MSI, and Razer.
Nvidia's GeForce RTX 5090 GPU arrives as the successor to the RTX 4090 model, and it is equipped with 32GB of GDDR7 VRAM (up from 24GB of GDDR6 VRAM), has a peak clock speed of 2.4GHz, and features 21,760 CUDA cores. The RTX 5080 model has 16GB of GDDR7 VRAM, and 10,752 CUDA cores.
Nvidia GeForce RTX 50 series compared with last year's RTX 4090 GPU
Photo Credit: Nvidia
On the other hand, the GeForce RTX 5070 Ti and RTX 5070 GPUs are equipped with 16GB and 12GB of GDDR7 VRAM, respectively. The former is equipped with 8,960 CUDA cores, while the latter has 6,144 CUDA cores.
All four GPUs are built using Nvidia's Blackwell architecture, which succeeds the Ada Lovelace technology. They feature fifth generation Tensor cores and fourth generation ray tracing cores, which are also a step up from the RTX 40 series models. Meanwhile, they are equipped with the ninth generation Nvidia Encoder (NVENC) and sixth generation Nvidia Decoder (NVDEC).
The GeForce RTX 5090 is the company's most powerful GPU for consumer PCs to date, and is equipped with 92 billion transistors, and the company says it offers up to two times the performance of last year's RTX 4090 model. Nvidia's Max-Q technology is also claimed to offer up to 40 percent improved battery life.
Catch the latest from the Consumer Electronics Show on Gadgets 360, at our CES 2026 hub.