From using GDDR6X memory to using GDDR6. The performance hit isn’t that much, but it is bullcrap that they did such a thing. The card is still named the RTX 4070. You have to be very careful when buying a 4070 now.
This is such a messed up change. Sure the performance hit is probably not noticeable in most cases, but the fact that they are downgrading a product without changing the name and causing confusion is horrible.
Unfortunately, with AMD saying they are not planning on competing with nvidia in the high end consumer GPU market, and the fact that Intel is currently on fire financially it leaves nvidia with no competition to stop them from doing things like this. Hopefully we will see ARM successfully compete with consumer GPUs as the rumors state, but I am not holding my breath.
I feel like AMD can actually pull ahead of Nvidia in the gaming space IF this new gen can compete with Nvidia, at a lower price. Most people get the budget xx60 or xx70 cards, not the xx80 or xx90. If they can release a xx60-xx70 level card at <=300$, it will be a smash hit most likely.
There are two challenges with this:
- AMD continues to not have a viable answer for DLSS upscaling. FSR upscaling is so completely inferior that they struggle to sell cards even when they have a small price advantage.
- They won’t sell them that cheap. AMD will probably continue to do what they have been doing for some time now. Launch their new GPUs at an outrageous price. Have all the reviewers point out how terrible they are. Then drop the price once they have missed the window for promoting the product.
That might come to an end soon.
That’s what I fear as well. They will keep doing this tactic that I’m sure ultimately works for them, but they really shouldn’t do because they leave out so much on the table.
I also fear that the next gen won’t be a massive improvement or that they can’t get a RTX 4060-4070 card at a decent price without skimping on something. We will have to see, though.