According to reports from Hong Kong media outlet HKEPC, Nvidia's highly anticipated RTX 5080 graphics card is expected to become available for purchase on January 21, following its anticipated reveal at CES 2024.
The launch timeline suggests Nvidia will officially unveil the RTX 5080 during CEO Jensen Huang's presentation on January 6, followed by a two-week sales embargo before retail availability. This pattern aligns with Nvidia's typical launch approach for new GPU releases.
Building anticipation for the announcement, Nvidia is reviving its GeForce LAN event after a 13-year hiatus, scheduled 48 hours before the CES reveal.
The RTX 5080 is reported to feature the GB203-400-A1 chip with 84 SMs delivering 10,752 CUDA cores - representing a 10% increase over its predecessor. The card will reportedly maintain 16GB of memory but upgrade to newer GDDR7 technology running at 30 Gbps, while other RTX 50 series cards are expected to use 28 Gbps memory.
Manufacturing improvements come from TSMC's custom 4NP (5nm-class) node, which offers 30% higher density compared to the previous generation's 4N process used in Ada Lovelace cards.
With AMD reportedly absent from the high-end GPU market this generation, the RTX 5080 may face limited competition, potentially giving Nvidia considerable pricing flexibility. However, final pricing details remain unconfirmed until the official announcement.
While HKEPC has previously provided accurate information, this launch date should be treated as speculative until officially confirmed by Nvidia. The timing of other RTX 50 series cards, including the RTX 5090 and RTX 5070, may differ from the RTX 5080's release schedule.