Nvidia may equip the GeForce RTX 4080 with higher-speed memory

Some time ago, it was reported that as the release time of the new generation Ada Lovelace architecture GPU is approaching, Nvidia adjusted the specifications of GeForce RTX 4070 / RTX 4080 again in the final stage. The power consumption of the two was reduced from the original 300W and 420W to 285W and 320W respectively after adjustment, which is much lower than previous rumors.
GeForce RTX 40 TSMC N5

However, such adjustments do not seem to satisfy NVIDIA completely. How to balance power consumption and performance to better counter AMD’s Radeon RX 7000 series based on the RDNA 3 architecture is an important consideration for NVIDIA. Recently, Twitter user @kopite7kimi revealed that NVIDIA has adjusted the GeForce RTX 4080 again, and the rate of GDDR6X memory will be increased from 21 Gbps to 23 Gbps, and the GeForce RTX 4090 is likely to have similar operations. Since Micron has recently started mass production of GDDR6X with a rate of 24 Gbps and a capacity of 16Gb, it is most likely to prepare for the GeForce RTX 40 series.

In addition, the power consumption of the GeForce RTX 4080 will be adjusted to 340W. Compared with the previous 320W, it is slightly improved, and it is possible that in order to adapt to the adjustment of the memory frequency, it is at the same level as the 350W of the existing GeForce RTX 3080.