Last night Nvidia put an end to several months of rumors by officially unveiling its new range of gaming graphics cards, the GeForce RXT 50 series. This family includes four models for this launch with the GeForce RTX 5090, RTX 5080, RTX 5070 Ti and RTX 5070. The recommended consumer prices range from $1,999 to $549.
The GeForce RTX 5090: the queen card
At the top of the range is the GeForce RTX 5090, featuring the GB202 GPU with 21,760 CUDA cores, a first for NVIDIA. Its equipment includes the presence of 32 GB of GDDR7 operated via a 512-bit bus, enough to ensure record bandwidth. The TDP reaches 575 W, a notable increase compared to the GeForce RTX 4090. Despite this consumption, NVIDIA has managed to design a Founders Edition in double slot format. The cooling system uses a 3D vapor chamber while the printed circuit board (PCB) is relatively compact. The GeForce RTX 5090 will be available for $1,999 starting January 30.
The GeForce RTX 5080: balanced performance
The GeForce RTX 5080, equipped with the GB203 GPU and 10,752 CUDA cores, has 16 GB of GDDR7 clocked at 30 Gbps. This is the highest speed available in the range. Its TDP is 360 W while price-wise it is positioned just under the $1000 ($999) mark. Its marketing is scheduled for the same day as that of the RTX 5090.
GeForce RTX 5070 Ti and RTX 5070: for more modest budgets
The GeForce RTX 5070 Ti uses 8,960 CUDA cores and is armed with 16 GB of GDDR7. Its price is $749. Finally, the GeForce RTX 5070 and its 6,144 CUDA cores accompanied by 12 GB of GDDR7, is announced as “twice as fast as the RTX 4070”. This model will be available for $549. Its launch is expected in February.
Note that Nvidia will not offer the RTX 5070 Ti in Founders Edition. This reference will be exclusively offered by AIB partners, while the RTX 5070 will benefit from a Founders Edition design. Finally, these new cards support PCIe 5.0, have GDDR7 and a DisplayPort 2.1a video output. They also feature a 16-pin power connector.
Follow Us
