We know that it is in the next few hours, at CES 2025 on January 6, that CEO Jensen Huang will present the new “Blackwell” architecture and several RTX 50s from this new range. So far, several hypotheses have emerged regarding the commercial launch of the cards after this event. HKEPC reported that the GeForce RTX 5080 will launch on January 21, 2025 and then removed its post from its social media channels. There RTX 5090 should follow and be on the market before the end of January according to the rumors we had until now.
RTX 5080, Intel B570, RDNA 4: A busy month of January
For the moment, it is difficult to comment on the positioning of these RTX 50s. We know from numerous leaks that the RTX 5080 will use a GB203-400 graphics chip. This will give access to a configuration of 10,752 CUDA cores which will be accompanied by 16 GB GDDR7 memory at 30 Gbps which, with a 256-bit memory interface, will provide a bandwidth of 960 GB/s. We mention a consumption of at least 400W which will justify the use of the new connector 12V-2×6 16 pin. How to situate this future RTX 5080 in relation to the current top of the range, the RTX 4090. Objectively it is difficult to say. The technical sheet of the RTX 4090 still seems superior to what the 5080 should offer. However, the supposed increase in the latter’s consumption leads us to believe that Nvidia will revise its operating frequencies upwards to get closer to it.
However, one thing seems to be clearly confirmed: the January calendar will be extremely busy. Intel should open the dance with the arrival of the B570 then Nvidia would therefore do this launch followed immediately by the arrival of AMD’s RDNA 4 cards. It is very possible that AMD’s schedule will be re-evaluated in this regard based on Nvidia’s announcements for the start of CES.
