Últimas Notícias

Nvidia plans RTX 50 Super launch with up to 24 GB in 2025

Nvidia
Nvidia - Foto: Hepha1st0s / Shutterstock.com Nvidia - Foto: Hepha1st0s / Shutterstock.com

Nvidia is gearing up to launch the highly anticipated RTX 50 Super GPUs, including the RTX 5070 Super, RTX 5070 Ti Super, and RTX 5080 Super, in the final quarter of 2025, capitalizing on the holiday season. According to rumors, these cards will feature a significant boost in VRAM, with up to 24 GB in top-tier models, and enhanced performance for gaming and AI applications. The information, yet to be confirmed by Nvidia, was reported by industry sources, pointing to an earlier-than-expected debut, originally slated for early 2026. The early launch aims to meet high demand from gamers and professionals, particularly during a peak commercial period. Nvidia’s strategy reflects the need to compete with rivals like AMD and deliver more robust options for 4K gaming and demanding tasks.

The RTX 50 series, based on the Blackwell architecture, debuted with the RTX 5090 and RTX 5080 in January 2025 at CES. Now, the arrival of the Super variants is set to bolster the company’s lineup with optimized specifications. These new cards are expected to attract both gamers and content creators seeking greater memory capacity and energy efficiency.

  • Key expected features:
    • 50% increase in VRAM capacity compared to base models.
    • Support for technologies like DLSS 4 and Multi Frame Generation.
    • Strategic launch timed for the holiday shopping season.

Expected specifications for the RTX 50 Super series

Rumors about the RTX 50 Super GPUs’ specifications suggest significant advancements, particularly in memory capacity and overall performance. The RTX 5070 Super is expected to feature 6,400 CUDA cores, a roughly 4% increase over the base model, while the RTX 5070 Ti Super may reach 8,960 cores. The RTX 5080 Super is likely to retain the 10,752 CUDA cores of the original but with a substantial VRAM boost. These details, though unconfirmed, were mentioned by reliable industry sources like leaker Kopite7kimi.

Nvidia’s big bet for this lineup is the use of 3 GB GDDR7 memory modules, enabling higher capacity configurations without increasing memory bus width. This results in greater efficiency for high-resolution gaming, like 4K, and data-intensive AI applications.

  • Specification details:
    • RTX 5070 Super: 18 GB GDDR7 VRAM, 275 W TGP.
    • RTX 5070 Ti Super: 24 GB GDDR7 VRAM, 350 W TGP.
    • RTX 5080 Super: 24 GB GDDR7 VRAM, 415 W TGP.
    • Support for DisplayPort 2.1b for high-resolution, high-refresh-rate monitors.

The increased power consumption (TGP) reflects the enhanced performance, but it also raises questions about the need for more robust power supplies. Nvidia appears focused on balancing performance and efficiency, particularly with technologies like DLSS 4, which improves visual quality with minimal performance impact.

Nvidia’s market strategy

The decision to bring forward the RTX 50 Super GPUs’ launch to late 2025, rather than waiting for CES 2026, underscores Nvidia’s aggressive strategy to maintain its lead in the graphics card market. The company faced criticism for limited supply during the RTX 30 and 40 series launches, leading to consumer frustration and inflated retail prices. With the Super series, Nvidia aims to avoid these issues, leveraging the holiday shopping season to maximize sales.

Competition with AMD, which is also planning to release competitive GPUs in the same period, is another factor likely influencing the decision. AMD’s RX 9000 series has garnered attention for offering strong performance at more affordable prices, pushing Nvidia to deliver products with greater value. The increased VRAM addresses a recurring demand from gamers and professionals who criticized the RTX 50 base models for memory limitations in demanding scenarios.

  • Factors driving the early launch:
    • High demand for GPUs with larger VRAM for 4K gaming and AI.
    • Intense competition with AMD in the high-performance segment.
    • Opportunity to capitalize on year-end sales.

Impact on gaming and AI markets

The RTX 50 Super GPUs are expected to solidify Nvidia’s position in the gaming market, particularly for titles leveraging advanced ray tracing and upscaling technologies, such as Cyberpunk 2077 and Alan Wake 2. The DLSS 4 technology, exclusive to the RTX 50 series, can generate up to three additional frames per rendered frame, boosting smoothness in high resolutions. This is especially beneficial for games with ray tracing in overdrive mode, which demand significant memory capacity.

Beyond gaming, the new cards target AI and content creation applications. The ability to run generative models locally with greater efficiency is a key selling point for creators using tools like NVIDIA ACE or Flux for image and video generation. The increased VRAM capacity supports processing large datasets, making the GPUs ideal for intensive creative workflows.

  • Benefits for different audiences:
    • Gamers: Enhanced performance in 4K gaming with ray tracing.
    • Creators: Support for AI and rendering workflows.
    • Professionals: Greater efficiency in machine learning tasks.

Performance and energy efficiency

The RTX 50 Super series promises performance gains, but the increased power consumption raises concerns. The RTX 5080 Super, with a 415 W TGP, requires robust power supplies, which could be a barrier for some consumers. Nvidia claims optimizations in power usage, such as Max-Q for laptops and FP4 support for AI models, help mitigate the impact. Still, users may need to invest in higher-capacity power supplies.

The Blackwell architecture, underpinning the RTX 50 series, brings significant improvements in computational efficiency, particularly with fifth-generation Tensor Cores and fourth-generation RT Cores. These optimizations enable the GPUs to handle complex workloads, like neural rendering and real-time ray tracing, more efficiently than the previous generation.

  • Highlighted technical aspects:
    • 5th-generation Tensor Cores for AI and neural rendering.
    • 4th-generation RT Cores for real-time ray tracing.
    • AV1 encoding support for high-quality streaming.

Consumer expectations

The initial reception of the RTX 50 series was lukewarm, with criticism over the base models’ performance gains compared to the previous generation, particularly the RTX 5080’s modest improvements over the RTX 4080 Super. The Super variants, with increased VRAM and optimizations, could shift this perception, but pricing will be a deciding factor. Nvidia has a history of launching Super models at competitive prices, but high demand and GDDR7 chip shortages may drive up retail costs.

Consumers hope the RTX 50 Super GPUs deliver a significant leap over the base models, especially for 4K gaming and AI applications. The prospect of a 24 GB VRAM card at a price close to the original RTX 5080 is seen as a major draw, but the lack of official pricing information keeps the market on edge.

  • What consumers expect:
    • Competitive pricing compared to base models.
    • Wide availability to avoid price gouging.
    • Robust support for gaming and AI applications.

Launch preparations

Nvidia has reportedly begun preparing for the RTX 50 Super launch, with units being stockpiled in warehouses to meet initial demand. The strategy includes partnerships with manufacturers like MSI and ASUS to ensure a diverse range of custom models. Additionally, Nvidia is investing in marketing campaigns to highlight the Super series’ advantages, particularly DLSS 4 and increased VRAM capacity.

The 2025 holiday season will be pivotal for Nvidia, which aims to reinforce its market leadership in an increasingly competitive landscape. The early launch could help capture consumer attention before CES 2026, but it also heightens pressure to deliver products that meet high expectations.

  • Launch measures:
    • Expanded inventory to prevent supply issues.
    • Partnerships with manufacturers for custom models.
    • Marketing campaigns focused on DLSS 4 and AI applications.
To Top