Here's the thing: while GDDR6X is capable of supporting a much higher bandwidth than GDDR6, actual performance depends on how the VRAM is clocked. The difference, in this case, may be negligible.

The latest VRAM standard may also be utilized in AI-powered workstations and non-gaming GPUs soon. On the technical side of things, first-generation GDDR7 supports a bandwidth of 32 Gbps per pin ...

TheMSI GeForce RTX 4080 SUPER 16G VENTUS 3X OC features16GB GDDR6X VRAM,2595 MHz boost clock,256-bit memory bus, and is powered by a single 16-pin connector.

As standard, the RTX 4070 comes with 12GB of GDDR6X VRAM running at 1,313MHz (21Gbps effective), giving you a total memory bandwidth of 504.2GB/s with this GPU's 192-bit memory interface.

But as RTX 4070 uses six VRAM chips, even if GDDR6 is only a few dollars per module cheaper, it could potentially mean such a variant could have a price tag lower by $30 or more.

For comparison, the GDDR6X in the current RTX 3080 runs at 19Gb/s, whereas GDDR6 can hit about 14Gb/s. It's possible Nvidia will compensate for this with a wider memory bus, or even extra L2 cache.

Read more: Chinese AI, cloud firms add VRAM to RTX 40 series GPUs: RTX 4090D with 48GB, RTX 4080 with 32GB A modded RTX 4090 with 48GB or the monster 96GB of GDDR6X memory won't be guaranteed with ...

With a switch from faster GDDR6X to slower GDDR6 memory, a new RTX 4070 variant could make for a slightly cheaper mid-range card. Edward Chester Published: 10 months ago Nvidia PC games hardware ...

Now, we've got a Chinese AI expert that is talking with people in the country about selling NVIDIA's new GeForce RTX 4090D graphics card -- but with double the VRAM, from 24GB to 48GB of GDDR6X ...

AMD’s Radeon RX 9000-series graphics cards feature16GB of GDDR6X VRAM. That’s still good — we can expect it to be quick, even if it’s not quite as fast as Nvidia’s latest cards.

However, it’s hard to compare: the price of GDDR6X from Micron is unknown. Since the RTX 4070 uses six VRAM chips, even if GDDR6 is only a few dollars cheaper per module, it potentially means a price ...