NVIDIA’s next-gen GeForce RTX 40 series graphics cards are set to launch in the latter half of this year, possibly as early as August. These GPUs will double the shader count from just 10,496 on the RTX 3090 to an incredible ~18,432 FP32 cores on the RTX 4090. There have been several reports claiming that these cutting-edge GPUs will be priced a fair bit higher than their predecessors. However, no concrete figures have surfaced to date.
|GPU||GA102||AD102||RTX 4090||AD103||RTX 4080||RTX 4070 Ti (AD104)||RTX 4070|
|Arch||Ampere||Ada Lovelace||Ada Lovelace||Ada Lovelace|
|Process||Sam 8nm LPP||TSMC 5nm||TSMC 5nm||TSMC 5nm|
|TP||37.6||~100 TFLOPs?||83 TFLOPs||~50 TFLOPs||47 TFLOPs?||~35 TFLOPs||35 TFLOPs?|
|Memory||24GB GDDR6X||48GB GDDR6X||24GB GDDR6X||16GB GDDR6X||12GB GDDR6X|
|Launch||Sep 2020||Sept 22?||Sept 22?||Q1 2023?|
A report from mydrivers states that the GeForce RTX 4090 will cost $1,899 at launch, a $400 increase over the RTX 3090 while essentially doubling the performance (tripling ray-tracing capabilities). This relatively mild increment in prices can be attributed to the improving condition of the global supply chain, increased GPU shipments, and falling prices.
Considering the higher cost of 5nm wafers from TSMC and GDDR6X memory, these prices aren’t too bad for enthusiasts and gamers, especially if you take into account the generational performance gains. The increased power consumption (up to 600W) also means these GPUs will be equipped with higher quality VRMs and components, again mandating an increase in both the BOM and MSRP.