R$ 36.999,99 for a graphics card. That was the Pix price of the Palit GameRock RTX 5090 at Pichau, checked on 18 October 2026, at around 12:13 p.m. Brasília time. The card payment price shown in the listing was R$ 43.529,40. The Gigabyte Gaming OC was listed at R$ 37.999,99 via Pix. These are two listings, without a shipping quote, rather than a survey of the lowest prices in Brazil. Even with that qualification, it is easy to see why following the battle at the top can feel like watching an entirely different hobby.
After all, who can put a 5090 in their cart at those prices? Buying one exclusively for gaming takes plenty of available money and a willingness to devote an enormous share of it to a PC. There is no obligation to make this market extreme the reference point for everyone. A generation can offer an appealing upgrade even when its most attention-grabbing product is completely out of the question.
The question that matters is closer to our own desks: what would a card available today add to the games we already have, and how much is getting that improvement now worth? Depending on the answer, your next upgrade might be in this generation. Or the best plan might be to keep playing while your money stays where it is.
The best card on the chart does not have to make your shortlist
One point of confusion is worth clearing up. In testing by Steven Walton, published by TechSpot on 14 October, the RTX 5090 reached 78 FPS against the RX 9070 XT’s 38 FPS across an average of 11 games at native 4K with more demanding settings: a 105% lead. At 1440p with lighter settings and Balanced upscaling, the figures were 224 versus 189 FPS, a 19% difference. The test system had a Ryzen 7 9800X3D, DDR5-6000 CL30, Windows 11 25H2, and AMD 26.8.1 and Nvidia 616.92 drivers.
Walton also recorded CPU limitations and the influence of driver overhead in specific cases. Those results do not make the Radeon faster overall, nor do they diagnose the reader’s computer. And the two scenarios cited change quality settings, resolution and reconstruction; they do not deliver the same image. The benchmarks are TechSpot’s, not MaisFPS’s.
For a purchase decision, what matters is noticing how the gap between products changes with their use. You can admire a card’s results and rule it out because of its cost. You do not have to prove that it is bad to conclude that it does not fit your life. The discussion about which model to buy starts after that filter.
Compare it with the card you already own
Reviews compare products on sale. Someone upgrading has another competitor: the installed GPU, which is already paid for. That is what the next purchase needs to prove itself against. A model can be competitive within its generation and still be a fairly unexciting replacement for the owner of a recent card.
Imagine two players, without tying them to specific models. One is already happy with the image and smoothness; the other has been skipping new releases because they cannot find settings that feel pleasant to play. Both come across the same deal. For the first, it buys refinement. For the second, it could make games in the library playable again. The price tag is identical; the value of the change is very different.
That is why it is worth judging the leap in terms of possibilities: using the monitor at the resolution you want, restoring details you miss, or comfortably playing through a campaign that has been waiting. An improvement you notice during play has a stronger justification than an advantage that only shows up when the counter is on.
This does not mean dismissing small improvements. Someone who plays for many hours may value them highly. The point is to recognize what you are buying, without letting the excitement of a comparison between generations inflate expectations for your own upgrade.
The game you want to play gives the decision a date
There is a difference between buying for next weekend and buying for a wish list with no deadline. If the campaign is installed, friends have agreed to start, and the current card is spoiling the experience, getting the improvement now has value. Delaying also has a cost, even when it does not appear on the bill.
A game you plan to try “when there is time” is a much weaker reason to buy ahead. Perhaps your tastes will change, perhaps the release will be delayed, or perhaps you will discover that you are still perfectly happy with the same old multiplayer game. Hardware bought for an imagined routine can spend quite a while waiting for its debut.
Resolution and visual preferences help make that plan concrete. Someone who wants to enjoy a detailed 4K screen is after a different experience from someone who prefers a competitive game with reduced visuals. Look for the games and settings that represent what you intend to do. A broad average is a starting point, not a description of your next month.
A new GPU can give the whole PC more time
Moving to a new graphics card generation does not mean replacing every component at once. If the platform still serves your games and the relevant limitation is graphics performance, keeping the rest can turn an unaffordable purchase into a possible improvement. The money stops funding a rebuild and goes where it makes a difference.
The important thing is to calculate the real package. The power supply, connectors and space in the case need to accommodate the chosen model. If the swap requires other spending, include it in the budget from the start. The benefit of reusing parts quickly disappears when each step reveals another purchase “just to complete the build”.
This also changes what useful life means. A computer stays useful as long as it does what its owner wants. It does not need to keep the same settings forever or run every new release at maximum settings. Accepting a visual reduction that barely bothers you can buy a lot of extra use; accepting an experience you no longer enjoy is another matter.
In our view, it is worth buying a comfortable margin for plausible needs. Paying an enormous sum to try to eliminate every future upgrade is a difficult bet to justify: no current benchmark guarantees what the games, demands and preferences of the coming years will be.
Waiting can be the right choice without predicting the future
You do not have to believe that the next generation will be wonderful to postpone a purchase. It is enough that current options do not offer an improvement worth the outlay. That is a conclusion about the present, which can be assessed using available listings and tests.
The next family’s price, its performance in your games and its availability in Brazil still need to exist before they can enter a serious comparison. Nor can anyone guarantee an automatic clearance sale of the previous generation. Waiting keeps your options open; it does not reserve an imaginary discount in your name.
A good way out of the anxiety is to decide what is missing for the purchase to make sense. It might be a bigger leap over the current GPU, a lower total cost or a need that has finally arrived. That lets you follow the market without having to take part in every launch. Reading about hardware can remain fun, without becoming an obligation.
So, this generation or another season with your card?
MaisFPS recommends buying now when the improvement is meaningful in the games you want, the full cost is comfortable and there is an immediate use for it. Keeping the current card makes sense when the experience is still enjoyable or the swap takes a lot of money to change very little.
The 5090 can remain an object of curiosity, desire or a choice for those who have the budget. Your next upgrade does not need permission from the top of the range. It needs to make your machine deliver something worth buying. If it already does, there is an excellent way to enjoy the current generation: close the store and open a game.
