Nvidia is about to launch the RTX 4070 Super GPU line

The latest leak says that the RTX 4070 Super GPU series, a mid-cycle refresh version of the RTX 40 “Ada” GPU series, will be sold by Nvidia in January 2024.

According to Wccftech, leaked information posted on a forum for Nvidia’s GPU partners in China said that Nvidia will introduce three GPU models of the RTX 4070 Super series on January 8, before the opening of CES 2024. just a day. Once introduced, Nvidia will offer the new GPU line to its partners, with the first product hitting shelves on January 17, followed by the RTX 4070 Ti Super on January 24 and the RTX 4080 Super on January 31.

Nvidia’s new GPUs are expected to deliver only modest performance gains over their predecessors, with the 4070 variants receiving more upgrades than the 4080 as they are more accessible to most people due to Lower prices and adequate resolution.

Regarding upgrades in the new Super GPU line, initial sources said Nvidia will bring a VRAM memory upgrade of up to 20 GB on the RTX 4080 Supermodel, but this is still a big question mark. Meanwhile, the RTX 4070 Ti Super is expected to bring a VRAM upgrade from 12 GB in the current card to 16 GB, which makes many gamers happy. It also provides a much-needed increase in memory bus levels, from 192 bits to 256 bits.

The RTX 4070 Super still has 12 GB of VRAM but gets more CUDA cores, from 5,888 to 7,168. Additionally, rumors suggest the card will use the new 12VHPWR connector – a big change as the current version uses the old 8-pin connector. Increasing the number of CUDA cores will also require an increase in TDP of 20W, bringing the total to 220W.

The success of these cards will depend on their price. Except for the RTX 4090, other members of the current RTX 40 GPU line are not too impressive in terms of power and performance but are sold at high prices, making gamers feel disappointed.

Related posts

GTA 6 is guaranteed to launch on time, Take-Two quashes delay rumors

Be wary of SteelFox malware attacking Windows using a copyright-cracking tool

Apple chose Foxconn and Lenovo to develop an AI server based on Apple Silicon