VRAM is a key component of what makes up a fast graphics card. It’s much faster than system memory and has much higher bandwidth, allowing for rapid access to things like game textures. It usually comes at a premium, with the last few generations of graphics cards having as high as 24GB and as low as 6GB.

The excuses have run dry as to why VRAM is so restricted on mid-tier cards. NVIDIA has restricted VRAM on its mid-tier offerings such as the 60-class to 8GB, with the price of grabbing the same card with more memory coming in at hundreds of dollars more. It’s getting a bit ridiculous, and it's definitely time for NVIDIA to start ponying up more VRAM on their mid-range offerings.

3 8GB is simply not enough these days

For how much these cards cost, we should be getting so much more

In mid-2023, the RTX 4060 and 4060 Ti were launched for $299 and $399, respectively. Not only were both of these cards launched with 8GB of RAM, they were preceded by a 12GB version of the 4060 Ti, paired with a $499 price tag. 8GB of VRAM is just not enough to reach the potential of the 4060 or the 4060 Ti, and having to pay at least $100 to have a card that's not kneecapped out of the box is a bit ridiculous. The 60-class cards routinely run into VRAM issues in AAA titles now, so it can only get worse if they continue to be stingy. That’s not the only weird behavior that NVIDIA engaged in during the 4000-series' tenure, however.

For those of you who may have forgotten, NVIDIA went as far as launching two different 4080 cards with 12 and 16GB of VRAM, and judging by name, you would think that there would be no difference in the computing ability of the actual chips. In actuality, these GPUs used completely different processing units and had very different levels of performance. The VRAM difference is significant on its own, and it’s definitely not worth the $300 price delta between the cards, even if they had the same processing unit.

2 VRAM does not add a significant amount of cost

NVIDIA certainly isn't pinching pennies, either

Adding a few more memory modules onto a GPU does not add hundreds of dollars of cost to the manufacturing of each card. I get that it’s high speed memory, and the bandwidth is high, but it would not kill them to equip the mid-range cards with a few more gigabytes of VRAM. Even AMD has been a bit more liberal with its implementation of VRAM on its mid-tier cards, so I’m sure NVIDIA can find enough change in the couch cushions to afford it.

1 The excuses for why aren’t valid anymore

The same old story just doesn't hold up now

It was often said that even if NVIDIA added the extra VRAM to their mid-tier cards, they wouldn’t be able to utilize it because of the lower bus width. This is nonsense, and was proven with AMD’s 7600 XT, which had 16GB of VRAM versus the 8GB offered on the regular 7600. The price difference between these two cards was only $60. A 2x increase in VRAM that costs you under $100. This is how you should be operating NVIDIA. It’s very easy not to gouge your customers for a few extra gigabytes of VRAM.

I hope there’s more competition on the horizon

With the launch of Intel’s ARC B-series GPUs being well received and new RDNA GPUs on the horizon from AMD, I sorely hope that there’s some more fierce competition in the mid-range GPU space. NVIDIA has gotten away with gouging for VRAM for far too long, and it's high time that more options are given to the consumer.