Nvidia has cancelled its planned January launch of the RTX 5080 24GB GPU due to soaring memory costs, according to a report from Board Channels.
High GDDR7 Chip Prices Halt Production Plans
The decision stems from the high expense of 3GB GDDR7 memory modules. These parts are reportedly three times more expensive than 2GB chips.
Moving to 3GB components would provide a 50% memory capacity upgrade over the 16GB variant. Overclock3D reports that the financial burden proved too difficult to justify for manufacturers.
These severe memory pricing pressures previously forced Nvidia to scrap its planned RTX 50 SUPER series of graphics processing units. Component costs have risen across the high-end hardware market.
Earlier reports pointed toward a potential introduction during CES 2027. Some sources indicated a release in the first quarter of the year.
Prior rumours suggested the 24GB variant might act as a new flagship replacing both the 16GB version and the RTX 5090 temporarily. It might also have launched under the RTX 5080 SUPER banner.
Nvidia has given no new release timeframe with the cancellation.
Board Partners Left Without Development Kits
Taiwanese media outlet BenchLife reported that add-in-board partners including ASUS and Gigabyte had not received the necessary design kits from Nvidia.
Partners were reportedly only told verbally that information regarding the hardware was on the way.
Nvidia has stopped supplying GB202 chips to board partners for GeForce RTX 5090 production. This restriction covers the China-specific RTX 5090 D and RTX 5090 D V2 models. These chips are being directed toward higher-margin RTX Pro cards instead. Retail prices are expected to climb further as available supply dwindles.
BenchLife describes this internal allocation change rather than a formal end-of-life procedure. What official statements will Nvidia share regarding the long-term outlook for the RTX 5090 line?
