Recently, there has been a resurgence of rumors concerning NVIDIA's upcoming RTX 50 Super series of graphics cards. The latest indication originates from an unexpected source - Seasonic's official power supply calculator. This tool now lists two unreleased GPU models, the GeForce RTX 5070 Super and RTX 5070 Ti Super, in its drop-down menu. While this doesn't confirm the imminent release of the Super series, it does suggest these products are undergoing manufacturer testing.

Speculations regarding the RTX 50 Super's specifications have been circulating from various sources. The primary upgrade in this series appears to be focused on memory capacity, with expectations of a significant 50% increase and the utilization of faster GDDR7 memory. For instance, the RTX 5070 Super is anticipated to feature 18GB of memory, while the RTX 5070 Ti Super could potentially offer up to 24GB. This enhancement is particularly crucial for the RTX 5070, which previously struggled with its 12GB capacity. The additional memory will likely provide better adaptability and an extended lifecycle for future high-resolution and graphically complex gaming environments.

In terms of specifications, both the RTX 5070 Super and RTX 5070 Ti Super are built on the Blackwell architecture. They offer CUDA core counts close to the non-Super versions, with 6,400 and 8,960 cores respectively, alongside improved memory setups and increased power consumption. Information leaked from the power calculator suggests these new cards have power ratings of 275W and 350W TDP, higher than the current RTX 5070 (250W) and RTX 5070 Ti (300W). Consequently, users planning to upgrade might require more powerful power supplies, but this comes with the benefit of enhanced video memory and possible performance expansion.

Beyond the 70-series, the Super Series also includes the RTX 5080 Super. This advancement maintains the core count at 10,752 CUDA cores, pairs with 24GB of GDDR7 memory, and offers 1024 GB/s bandwidth. It's expected to exceed 400W power consumption, in contrast to the regular RTX 5080, which features 16GB memory, 960 GB/s bandwidth, and a 360W TDP. These changes align with the 70-series strategy that primarily focuses on memory expansion and elevated power usage while maintaining consistent architecture and core size.

It's important to note that NVIDIA has yet to confirm the Super Series officially, and its release timeline remains uncertain. Previously, it was reported that NVIDIA would not unveil any Super models at CES 2025, yet Seasonic's tool update could indicate that manufacturers are already preparing for these potential products. Even if the Super lineup isn't immediately launched, NVIDIA might adjust its release schedule in future quarters based on market needs and competition.

Should the RTX 50 Super series be realized, its market positioning will be clearly defined: slotting between existing models in price and performance. This series would offer high-end gamers enhanced memory and bandwidth options to meet the demands of next-generation gaming and AI applications. In response to AMD's and Intel's accelerated moves in the GPU market, NVIDIA's strategy to diversify with the Super Series could be seen as a logical market adaptation.