CES 2026 will ultimately provide answers to significant questions regarding Nvidia's RTX 50 Super GPUs.

      With Nvidia's Blackwell architecture firmly established in AI data centers, cloud services, workstations, and desktop/laptop PCs, the company’s CES 2026 press event is anticipated to concentrate less on new releases. Instead, it is expected that Nvidia will explore enhancements, future plans, and its strategy for advancing its hardware and software stack in the upcoming year.

      Nvidia has announced that founder and CEO Jensen Huang will present the company's opening keynote at CES on January 5, a day before the main CES exhibition begins. CES has evolved into one of Nvidia's most significant platforms annually, even in years when the company is not unveiling a new generation of GPUs.

      As has been the case in previous years, Nvidia is likely to highlight its key areas of focus, including gaming, AI, data centers, and emerging technologies. Additionally, the company will have a strong presence at the event with possible demonstrations and showcases.

      Here’s a brief overview of what Nvidia might present next month.

      RTX 50 Super

      If there are any consumer GPU announcements from Nvidia at CES 2026, they will probably pertain to the RTX 50 Super. According to leaks and reports, the Super refresh was initially targeted for a late 2025 launch to coincide with the holiday season.

      However, due to increasing memory costs and Nvidia’s history of extending refresh timelines, a postponement to early or mid-2026 now seems more plausible, with CES serving more as a platform for showcasing rather than launching.

      The RTX 50 Super is not anticipated to bring significant changes. Rather, it appears to be a minor update to the current Blackwell GPUs, featuring adjusted specifications aimed at enhancing the balance of the lineup. The RTX 40 Super refresh followed a similar trend, and current leaks indicate that Nvidia is adopting the same strategy.

      From the information available so far, Nvidia is likely to target specific SKUs instead of an entire lineup refresh. Rumored models include the RTX 5070 Super, RTX 5070 Ti Super, and RTX 5080 Super. The RTX 5070 Super is expected to shift to a higher VRAM configuration, potentially 18GB, addressing a major complaint regarding the standard model.

      The RTX 5080 Super is rumored to maintain a similar core layout but could feature 8GB more memory for a total of 24GB, along with enhanced bandwidth, potentially bringing its performance closer to the RTX 5090 in practical workloads rather than just raw computing power.

      Graphics card RTX 5080 Super RTX 5070 Ti Super RTX 5070 Super

      Architecture Blackwell GB203 Blackwell GB203 Blackwell GB205

      VRAM 24GB GDDR7 24GB GDDR7 18GB GDDR7

      VRAM Bus Width 256-bit 256-bit 192-bit

      CUDA Cores 10,752 8,960 6,400

      TGP 415W 350W 275W

      Any performance improvements are expected to be moderate. Small clock increases and minor core modifications are more probable than substantial rises in CUDA counts. The aim seems to be to reduce gaps within the product lineup rather than making a distinct generational leap.

      Memory configuration is a key reason for any refresh at all. As games and software increasingly require more memory, some RTX 50 models are already facing limitations. A Super refresh would provide Nvidia with the opportunity to update these configurations without altering the fundamental architecture.

      The timing, however, remains uncertain. Rising memory prices complicate changes in configurations without increasing costs, giving Nvidia a clear motive to wait. Consequently, CES 2026 is more likely to offer confirmation rather than immediate availability.

      Nvidia may discuss the RTX 50 Super or outline its plans, while reserving the actual release for later in 2026. This clarity could help buyers decide whether to upgrade now or hold off a bit longer.

      AI to take center stage

      While gamers are concerned about GPUs, AI will remain Nvidia’s primary focus at CES 2026. Anticipate the company to dedicate considerable time discussing AI hardware for enterprises and data centers, particularly accelerators designed for inference, training, and edge deployment.

      Though CES isn’t GTC, it has increasingly become a platform for Nvidia to demonstrate how its AI stack scales from large data centers to local AI PCs.

      Automotive technology is another likely focus area. Nvidia has a long history in autonomous driving and in-car AI systems, and CES is typically where automotive manufacturers and chip producers align their strategic plans. New announcements regarding DRIVE platforms, in-car AI computation, or expanded partnerships would not be unexpected.

      In addition to automotive, Nvidia might also address embedded systems, industrial AI, robotics, and edge computing. These sectors are where Nvidia’s hardware and software frameworks already exhibit dominance, and CES offers the perfect setting to illustrate how this technology enhances everyday products.

      The Arm wild card

      One of the most intriguing