His love of PCs and their components was born out of trying to squeeze every ounce of performance out of the family computer. Tinkering with his own build at age 10 turned into building PCs for friends and family, fostering a passion that would ultimately take shape as a career path.
Besides being the first call for tech support for those close to him, Ty is a computer science student, with his focus being cloud computing and networking. He also competed in semi-pro Counter-Strike for 8 years, making him intimately familiar with everything to do with peripherals.
The RTX 5090 is the most powerful gaming GPU ever released, and it comes with 21,760 CUDA cores on its GB202 Blackwell chip, which is just over two times the amount that comes on the RTX 5080. If October's leaks are correct and a 24GB RTX 5080 becomes the best GeForce card that gamers can buy, Nvidia's flagship gaming offering essentially loses half of its compute overnight. Jeff Fisher, Senior VP of everything GeForce at Nvidia, has told Taiwanese media that the 5090 isn't going anywhere, but that doesn't mean much when nobody can actually buy one.
A 24GB RTX 5080 is still half of an RTX 5090
The rumored launch of the 24GB variant of the RTX 5080 is expected to carry the "Super" branding that has been dormant since the 40-series GPUs, but this new variant isn't changing anything about the main silicon. The RTX 5080 already uses the full GB203 die, so Nvidia has nothing left to unlock without moving to a cut-down GB202, which is already in short supply. The leaks also describe a power bump, going to a 415W TDP, a 55W increase over the 360W on the standard card.
For those who game at 4K with the graphical fidelity cranked, the memory and power upgrades are meaningful. 16GB already causes trouble in the most demanding games. Indiana Jones and the Great Circle, for example, breaks down at 4K with path tracing on a 16GB card. Even so, a memory upgrade on a card that launched in January 2025 is not a new flagship, and it just means Nvidia would be topping its GeForce lineup with silicon it has been selling for almost two years already.
Realistically, the market has already made this choice, because the 5090 was unattainable for the vast majority of gamers, even those that could afford the $1,999 MSRP price tag. The cheapest RTX 5090 is currently around $8,000 at the time of writing, while the cheapest RTX 5080 is around $1,700. The 5080 has been Nvidia's real top gaming card for awhile now.
The 80-class has been drifting away from the top for three generations
The gap has been getting larger
Two generations ago, the gap looked pretty different: the RTX 3080 had 8,704 CUDA cores to the RTX 3090’s 10,496, or about 83%. Both cards came from the same GA102 die, so the 3090 was essentially a "fuller" version of the 3080, though they were much closer than the cards that came after them.
The RTX 4080 moved to the smaller AD103 die while the 4090 stayed on AD102, and the 4080’s 9,728 cores came to about 59% of the 4090’s 16,384, and that continued with Blackwell. Over three generations, the 80-class cards went from being most of a flagship to being about half of one.
As someone who owns both an RTX 3080 and RTX 5080, I have never felt like either wasn't enough for high-resolution gaming for the titles released around the time of the card. The 3080 was especially an excellent value in retrospect, considering it can still handle many newer games at 4K with Quality upscaling. The 5080 is definitely a meaningful upgrade, providing much better 1% lows over the 3080 when the resolution isn't as high, but with its bigger VRAM buffer and bump in compute helps it be an everyday 4K gaming card without upscaling.
Compared to what the alleged RTX 5080 Super will be, it's hard to tell without having one in-hand, but I haven't come across a game personally where the 16GB capacity was a limiting factor in performance, even with texture quality and filtering maxed.
The GB202 has more profitable places to be
The 5090 turned out to be a much better compute card than a gaming one
GB202 powers both the RTX 5090 and the RTX Pro 6000 Blackwell workstation card. Every GB202 that becomes a GeForce card is one that can’t become a much more expensive Pro card, and it's no secret that even the RTX 5090 is highly sought after for its prowess as a workstation card over a gaming one.
Fisher’s denial addressed end-of-life, not how GB202 chips are allocated, and the popularity of the silicon for the workstation purpose being far more profitable means that the price of the 5090 is not coming down anytime soon. It doesn't mean the leak is true, and Nvidia could simply be struggling to source enough GDDR7 for a 32GB consumer card. Buyers don't care about that, though; if stock doesn't exist for them to buy, the result is the same either way.
We'll find out more about the alleged 50-series Super refresh soon enough
CES 2027 should be the next checkpoint for the GeForce brand as a whole. If Nvidia announces a 24GB RTX 5080 there, how it’s positioned matters a lot: if it's the same GB203 with more memory, presented as the best GeForce card you can actually buy, the GB202 is essentially dead for gamers. There are rumors of an "SE" variant of the 5090, which includes a cut-down GB202, but those rumors are unsubstantiated at the time of writing.