The Nvidia GeForce RTX 4080 is reportedly dead, and by Geralt I hope this latest round of tech gossip is true. The RTX 4080 has been a clown car on a PCB since it was unleashed at a completely baffling price of $1,199 at the end of 2022. If graphics cards had feelings, I’d feel a bit sorry for it and give it a caring pat on its oversized cooler. The RTX 4080 isn’t a bad GPU, after all, it just has a price that was seemingly conjured up during a ‘pin the GPU on the price chart’ party game.
Somehow, this doomed Nvidia GPU even managed to get off to a bad start before the first GeForce RTX 4080 cards even landed on the shelves. Cast your mind back to last year, and you may remember there were originally two RTX 4080 cards slated – a cheaper GeForce RTX 4080 12GB card, and a pricier 16GB version, both with vastly different GPU specs beyond the memory allocation.
People pointed and laughed at them, and Nvidia eventually raised a white flag and renamed the RTX 4080 12GB the GeForce RTX 4070 Ti, a GPU that’s also overpriced, but not to quite the same comical degree as the RTX 4080.
Now, however, the GeForce RTX 4080 is being discontinued, apparently, along with the GeForce RTX TX 4070 Ti, in order to make way for a new series of GeForce RTX 4000 Super graphics cards next year. Not only that, but the new RTX 4080 Super price is rumored to be under $1,000. That’s still a ridiculous price, of course, but if these rumors are true, they show Nvidia rightfully reflecting on what went so wrong with the RTX 4080.
And what went so wrong was mainly the price. Nvidia has been using the same numbering series for its GPUs for over a decade now, and in that time the 80-numbered GPU in each lineup has gradually increased in price. As you can see in the graph below, the launch price of $499 for the GeForce GTX 480 took seven GPU numbering generations to get to $699 for the GeForce RTX 3080, stopping at $549 and $599 on the way. Then, suddenly, the price rockets by $500 for the GeForce RTX 4080.
There are a few factors at play here, of course. For a start, the $699 price for the RTX 3080 was largely fictional in reality, because the pandemic and a GPU crypto-mining boom resulted in supply that couldn’t meet demand. Scalpers swept up any gaming stock as soon as it appeared, and put the cards straight up on eBay for comically inflated prices. What’s more, people bought them. There’s also been a silicon shortage and global inflation bumping up prices further. But even so, $1,199? Really? No, even accounting for these factors, the RTX 4080 does offer extraordinarily bad value, it’s not your imagination.
Just to make the point about value, let’s make a crude calculation of bang per buck, based on CUDA cores per dollar at the launch price of all the GPUs in the Nvidia Ada lineup. Real life isn’t this simple, of course – there are other factors such as memory, clock speed, and bus width that make big differences to graphics card performance too. However, this rough calculation does demonstrate the shockingly poor value of the RTX 4080.
There are some small fluctuations in the graph above, with the RTX 4070 Ti representing worse value than all the GPUs below it, but you otherwise get 9.8-10.9 CUDA cores per dollar across the range, apart from the RTX 4080, which only buys you 8.1 CUDA cores per dollar. Even the expensive RTX 4090 offers substantially better value than the RTX 4080 – the price of the RTX 4080 simply doesn’t bear any relation to reality.
Off the record, more than one retailer has told me that sales of GeForce RTX 4080 cards have been disappointing, and I’m entirely unsurprised. It’s fine to have a flagship product that costs a lot of money – that’s the card for people who want the very best and are happy to pay for it. Those people will buy the RTX 4090, though. Why on earth would someone with over a grand to spend on a GPU buy an RTX 4080 when the RTX 4090 is so much better?
For everyone who isn’t in a position to buy the very best, you need your product stack to look tempting enough for someone to part with that price. I’m usually the target market for an Nvidia GPU with an ‘80’ in its name (I’ve had a 680, a 780, a 980, a 1080, and a 3080), but I’ve passed on the RTX 4080 completely.
Had it launched at $749 with 10,752 cores I’d have bought it on launch day, as I said when looking at what the Nvidia Ada lineup should have been. I’d have seriously considered it at $100 more than that, but there’s just nothing tempting about a graphics card that offers such shockingly poor value.
I’m not expecting the new Super range, if it does indeed come out next year, to offer amazing value, based on the current speculation, but let’s at least hope the price of the new GeForce RTX 4080 Super at least makes some sort of sense.
Did you buy a GeForce RTX 4080? Let us know your thoughts and questions on this post by joining the conversation on the CustomC Facebook page, via Twitter, or join our Custom PC and Gaming Setup Facebook group and tap into the knowledge of our 420,000+ members.