The RTX 4080 was to be offered in two variants – one with 9,728 CUDA cores and 16GB of GDDR6X memory and a lower-tiered spec featuring 7,680 CUDA cores alongside 12GB of Micron GDDR6X memory – and launch in the middle of October starting at $1,199 and $899, respectively. Public outcry regarding the lower-spec card prompted a rethink from Nvidia, and in November the company “unlaunched” the 12GB RTX 4080 before it was ever released because it “wasn’t named right.” That brings us to today and the formal introduction of the RTX 4070 Ti.
Nvidia’s super misleading graph: up to three times the performance ** but comparing regular rendering vs. DLSS 3. No clear footnotes to indicate such though. According to Nvidia’s product page, the RTX 4070 Ti is faster than the GeForce RTX 3090 Ti at nearly half the power. It will “max out your 1440p monitor” with frame rates north of 120 FPS in modern titles like Warhammer 40,000: Darktide and A Plague Tale: Requiem The card packs 7,680 CUDA cores and has 12GB of GDDR6X memory with a base clock of 2.31GHz and a boost clock of 2.61GHz. It also boasts a 192-bit memory interface. On average, you can expect the card to draw around 226W of power. No surprises here – it’s simply a rebranded RTX 4080 12GB. What has changed is the cost – it now starts at $799. It’ll be available starting January 5 from manufacturing partners including Asus, Gainward, Gigabyte, MSI, Palit, PNY and Zotac, among others. It’s unclear what retail availability will look like but given how much extra time Nvidia has had to prep the card, we can only hope there will be sufficient inventory to meet initial demand.
