NVIDIA has released a new driver that removes the ability to modify the power limit on RTX 5000 series graphics cards. This feature, known as Power Limit , allowed users to manually adjust the maximum power the GPU could consume, a common option for overclocking and optimizing performance under load. The removal has surprised the community, as the company has not offered an official explanation.
The change comes at a time when NVIDIA is reorganizing its product lineup and preparing for new generations. According to sources close to the company, production of the GeForce RTX 5070 Ti has already ended, and the next significant model would be the 16GB RTX 5060 Ti. Furthermore, the company is rumored to be working on a potential RTX 5090 SE and an RTX 50 SUPER series for CES 2026, while the Rubin architecture is expected to power the future RTX 60 series in 2026.
A change in the driver that affects overclocking
Disabling the Power Limit on the RTX 5000 series means users lose a key manual adjustment method for controlling power consumption and temperature . Until now, this option allowed users to raise or lower the power limit to maximize the card's performance or reduce heat generation. Without it, overclocking options are limited, and many enthusiasts have expressed their discontent on specialized forums.
NVIDIA has not clarified whether this is a technical fix, a support decision, or a measure to prevent stability issues. It also hasn't indicated whether there will be an alternative way to restore this functionality in future driver versions. In the meantime, users who relied on this setting will have to make do with the factory defaults or resort to third-party tools, although these may not be compatible.
RTX 50 Series Context
The GeForce RTX 50 series, unveiled at CES 2025, is based on the Blackwell architecture and represents a generational leap in performance and efficiency. It includes models such as the RTX 5090, RTX 5080, RTX 5070 Ti, RTX 5070, RTX 5060 Ti, RTX 5060, and RTX 5050, all featuring GDDR7 memory (except for the RTX 5050, which uses GDDR6) and PCIe 5.0 support. The high-end RTX 5090 stands out with its 21.760 CUDA cores, while the RTX 5070 promises performance similar to the RTX 4090 thanks to DLSS 4 and multi-frame generation.
However, the series has not been without controversy. Since its launch, there have been reports of availability issues and prices well above the MSRP , especially in Europe. Furthermore, some users have experienced problems such as melting 12V-2x6 connectors, black screens after installing drivers, and questionable driver quality in the 572.XX and 576.XX branches. These problems have affected not only the RTX 50 series but also previous generations like the RTX 30 and RTX 40 series.
Problems and criticisms
Criticism of NVIDIA has focused on performance claims that don't hold true in practice. For example, the RTX 5070 doesn't match the RTX 4090's performance in real-world games unless DLSS 4 with multi-frame generation is enabled, which introduces latency and visual artifacts. The RTX 5060 Ti, meanwhile, has been criticized for its limited VRAM in the 8GB version, which isn't sufficient for high resolutions. Even the top-of-the-line RTX 5090 has received a lukewarm reception due to its high price and its reliance on AI to justify its value.
Another point of contention has been NVIDIA's decision not to send samples to the media for review of certain models, such as the RTX 5060 Ti 8GB and the RTX 5050. This has generated distrust and led many analysts to question the company's transparency. Furthermore, the drop in support for 32-bit PhysX and its subsequent reinstatement in December 2025 has added fuel to the fire.
Future of the range
Meanwhile, NVIDIA is already looking ahead. According to reports, production of the RTX 5070 Ti has ceased, and the 16GB RTX 5060 Ti will become the flagship model in the mid-range. There's also talk of a possible RTX 5090 SE, a cut-down version of the top-of-the-line card, and an update to the series with RTX 50 SUPER GPUs that could be unveiled at CES 2026. Further down the line, the Rubin architecture is expected to power the RTX 60 series, which is anticipated for 2026.
In the market, NVIDIA continues to dominate with a 94% share of the discrete GPU market, according to the data collected. This dominance is further reinforced by the upcoming integration of RTX iGPUs linked to Intel, which would expand its ecosystem even further. However, the company will need to resolve current driver issues and address the negative perceptions of some users if it wants to maintain its leadership.
In short, disabling the Power Limit on the RTX 5000 series is a move that affects overclocking enthusiasts, but it's part of a broader context of changes and challenges for NVIDIA. The RTX 50 series remains powerful, but its launch has been marked by controversy. With new products on the horizon and an overwhelming market share, the company has an opportunity to turn things around, although for now, users will have to adapt to the new software restrictions.