Recently, we reported that Nvidia has likely canceled the new GeForce RTX 5000 Super graphics cards planned for this year, and that the wait for the 6000 generation will also be long, as it is expected to arrive only in H2 2027. For gamers, the situation may get even worse, as the company is now reportedly reducing production of currently available graphics cards as well. A resulting shortage could then push retail prices higher.
Fresh information surfaced online yesterday from a fairly reliable leaker in China known as MEGAsizeGPU. According to his sources, Nvidia has reduced the number of GPU chips it supplies to graphics card manufacturers by 15 to 20 % compared to previous levels. This effectively means fewer graphics cards—namely Nvidia GeForce models—will be shipped to the market.
It is not possible to determine whether a similar cut also applies to Founders Edition cards sold directly by Nvidia, but models from third-party manufacturers account for the majority of the market, so the impact should be tangible. Smaller GPU allocations are likely to worsen availability, raising the risk of price increases—potentially severe ones. This is already visible with the GeForce RTX 5090, whose prices have surged dramatically compared to its original official MSRP of 1999 USD and are now reaching close to 4000 USD globally.
One piece of relatively good news amid these developments is that, according to MEGAsizeGPU, Nvidia is still bundling memory with GPU shipments to card manufacturers. This is critical, as these are smaller companies that would be unable to independently secure memory chips at even remotely reasonable prices under current market conditions.
This arrangement likely helps slow further price increases that would otherwise occur, but it is also probably one of the reasons Nvidia has reduced the volume of GPUs supplied to partners. Earlier reports suggested that Nvidia would stop bundling memory with GPUs, which—according to MEGAsizeGPU—has not happened yet, or at least not for the partner manufacturers from whom they have information.
Author: Wirestock Creators / Shutterstock
MEGAsizeGPU also corroborates reports that no new Nvidia graphics cards will be released this year—meaning no Super models—and the claims that the GeForce RTX 6000 generation is still far off. It has either been delayed somewhat or Nvidia was always planning for the GeForce RTX 5000 series to be followed by a longer gap after than was customary in the past. Per MEGAsizeGPU, GeForce RTX 6000 GPUs will launch in neither the first nor second quarter of 2027, and little actual information about them exists so far. The release of this new Rubin-based generation therefore now looks more likely to be slated for summer or fall 2027, meaning the previously regular two-year cadence—where a new generation arrived every even-numbered year—will slip by a full year. RTX 6000 would then arrive five years after the debut of the GeForce RTX 4000 generation.
- Tip: GeForce RTX 5000 Super not coming, Blackwell refresh plan shelved
- Tip: GeForce RTX 6000 delayed, launch at least a year and a half away
GeForce RTX 5060 Ti 16 GB and GeForce RTX 5070 Ti 16 GB will disappear
According to reports from additional sources, the production cuts will not affect all GeForce models equally, but will instead target certain SKUs. Nvidia is reportedly prioritizing the highest-end and most expensive card variants for each given memory configuration, reflecting the fact that GDDR7 availability is the main limiting factor. In practice, this means that among cards with 8 GB of GDDR7, the GeForce RTX 5060 Ti is being prioritized at the expense of the cheaper RTX 5060, while among 16 GB models, production favors the GeForce RTX 5080.
The GeForce RTX 5070 with 12 GB of memory is not mentioned in these reports, but it is likely strategically important for Nvidia’s marketshare dominance, so its cancellation would be surprising. However, the GeForce RTX 5070 Ti appears to be more or less vacating the market. Asus has reportedly told the Hardware Unboxed editorial team that GPU chips for this card are no longer available and that production will not continue. The reason is straightforward: the same number of GDDR7 memory chips can be used to produce a more profitable GeForce RTX 5080, which Nvidia—and likely board partners as well—prefer to sell (if the customer demand allows).
For the same reason, GeForce RTX 5060 Ti models with 16 GB of memory are also expected to disappear or greatly diminish in retail. Again, this is because the memory required to produce such cards could instead be used for higher-margin products like the RTX 5080—or alternatively for two RTX 5060 Ti 8 GB cards.
Nvidia has responded to these reports—which is somewhat unusual, as the company typically avoids commenting on rumors—telling Hardwareluxx that it has not discontinued production of any GPU variant and continues to supply all models. However, Nvidia did not specify whether shipment volumes remain unchanged, so it is quite possible that while the GeForce RTX 5060 Ti 16 GB and GeForce RTX 5070 Ti still exist on paper, manufacturers could be receiving them in significantly reduced quantities.
Reports of these models being constrained and of higher-end variants being favored have emerged from multiple sources, and information from Asus in particular supports the idea of some form of production volume cuts. Asus has since issued a statement clarifying that it has not canceled or stopped producing the GeForce RTX 5070 Ti and GeForce RTX 5060 Ti 16 GB. According to the company, the limited availability of these SKUs is due to memory supply issues. This aligns with the interpretation that these products are being deprioritized and throttled rather than formally discontinued, with shipments being dialed down instead of the card being discontinued.
Sources: VideoCardz (1, 2, 3), Hardware Unboxed, MEGAsizeGPU (1, 2)
English translation and edit by Jozef Dudáš
⠀
