NVIDIA Reveals Detailed Specs for GeForce RTX 5070 Ti & RTX 5070 with “Blackwell” Chips

NVIDIA has finally lifted the curtain on the highly anticipated GeForce RTX 5070 Ti and RTX 5070 graphics cards, unveiling their full specifications within the powerful Blackwell lineup. Dubbed the next generation of gaming GPUs, these cards promise significant improvements in power and performance.

Starting with the GeForce RTX 5070 Ti, this graphics card is now officially positioned as the third-fastest in the series, taking over from the previous RTX 4070 Ti. Under the hood is the robust GB203 GPU, which shares some core similarities with the RTX 5080, including a 16 GB GDDR7 VRAM. However, the RTX 5070 Ti offers slightly reduced memory speeds of 28 Gbps in comparison to the RTX 5080’s 30 Gbps.

Meanwhile, the GeForce RTX 5070 is built around the GB205 GPU die, setting it as the slowest among the newly revealed Blackwell GPUs. Despite this, NVIDIA has packed it with notable enhancements, such as increased CUDA Cores, Tensor Cores, and RT Cores, promising a solid performance bump. While maintaining a 12 GB GDDR7 VRAM on a 192-bit bus, this model slightly edges past the RTX 4070 with a boost clock of 2512 MHz — an improvement over its predecessor’s 2475 MHz.

Energy consumption is an important factor, with the RTX 5070 Ti consuming around 300W, and the standard RTX 5070 drawing 250W in the Founder’s Edition model. Of course, other versions with custom cooling solutions might vary in power usage. Expected launch dates have the RTX 5070 Ti hitting the market on February 20, while the RTX 5070 will follow in March. Both are set to debut with a price cut of $50 compared to previous iterations, though premium custom editions might command higher prices.

Delving deeper into their capabilities, the RTX 5070 Ti impresses with its 8960 CUDA Cores and elevated memory bandwidth of 896 GB/s. Performance jumps to double that of the RTX 4070 Ti, all for a price of $749, with cutting-edge DLSS 4 features adding to its appeal.

The RTX 5070, described as a formidable contender against higher-end GPUs, offers 6144 CUDA cores and enhanced memory speeds that achieve a 33% bandwidth improvement compared to its predecessor. Priced attractively at $549, it emerges as a strong competitor in the high-performance sector, challenging rival AMD’s RX 9070 series.

Both cards are designed optimally for 1440p gaming, while enthusiasts looking for 4K performance will find options like the RTX 5080 more suitable. NVIDIA continues to expand its groundbreaking RTX 50 series, paving new pathways in gaming technology and offering versatile options to cater to a wide range of gamers’ needs and budgets. Keep an eye out as these innovations hit the shelves, revolutionizing the gaming landscape once more.