Why the RTX 50 series is the most disappointing NVIDIA GPU generation

The latest GPU structure in NVIDIA, Blackwell, is impressive when it comes to the company’s data center. But on the side of the games from the street, things are not pink. In fact, RTX 50 graphics processing units are so disappointing that I consider RTX 50 the most disappointing Nvidia GPU generation never.

Different performance of hopes compared to the 40 series

NVIDIA has always delivered goods related to raising performance on the previous GPU. For decades, all new GTX, and RTX later, the collection of the previous generation graphics card will explode.

We had some appropriate Duds that barely transferred the game needle, such as the GeForce 9000 series. But that was a simple update of the Tesla architecture that first appeared on the GTX 8000 series. Regardless of the GTX 9800 GTX and 9800 GTX+, GeForce 9000 graphics processing units were the same as in terms of performing games compared to GTX 8000 counterparts.

Even the world’s frightening generations, such as RTX 20, were delivered as it was important. With RTX 20, we got DLSS, and although the RTX 2070 did not exceed the Halo product for the previous generation, the legendary GTX 1080 TI, the RTX 2080 easily jumped on the previous performance in the performance of the games.

Then the RTX 40 series came, and there was something. Although the RTX 4090 was a monster, other family members offered a somewhat disappointing performance that reminds us of the RTX 20 series while huge prices rocked compared to RTX 30 bonuses.

If the RTX 40 series is disappointment, the RTX 50 collection is a flat disaster with regard to improving performance on the previous generation. The only luminous spot is RTX 5090, with the rest of the stack that barely rises on their predecessors.

RTX 5080 only exceeds about 12-15 % of RTX 4080 and miles behind RTX 4090, something that no one has heard in the past when the XX80-Series graphics unit has always overpowered the previous generation. RTX 5070 Ti RTX 4070 TI exceeds 20 %.

The RTX 5070, which comes with 23 to 25 % solid performance on its predecessor-but is still seen on the RTX 4070 Super as a disappointment thanks to the misleading NVIDIA marketing that claims to be claimed to be RTX 5090.

Availability problems are still standing

Inability to find new gaming graphics processing units at all is completely normal. Even a decade ago, you will have difficulty picking up the graphics processing unit a few weeks after its appearance. But about a month and a half after its landing, the RTX 5090 is still in the event of a deficiency in the width. If you find it in one way or another, be ready to spend the equivalent of a decent used car.

RTX 5080, which is launched on the same day as RTX 5090, is intermittent in stock, both on Newgg and Best Buy. But you can only get expensive OC models, which are at least a few hundred dollars from the MSRP card for $ 1,000.

RTX 5070 TI was launched just one week after the two main options, but it is also rare to be found in the wild. And if you can locate one location, it will help you again, pay a muffled bonus on MSRP at a value of $ 750. Finally, RTX 5070 was released just a few weeks ago. Therefore, although it is normal not to find it in the stock, I think it will also remain difficult to find in the coming weeks.

Sales data confirm that it is difficult to obtain RTX 50 cards. According to Nafidia, Via the edgeThe RTX 50 Series are shipped twice the number of units compared to the RTX 40 series within the first five weeks of availability, but there is hunting.

While we have seen the launch operations of four SKU RTX 50 SKU over the past five weeks, the only RTX 40 graphics processing unit available during the same period in 2022 Halo, RTX 4090.

In short, NVIDIA claims that it took four RTX 50 models to double charging numbers in RTX 4090, which really did not fly from the shelves due to MSRP worth $ 1,600. At least, you can get the RTX 4090 in or near MSRP for a period of time. Good luck enables any RTX 50 cards in MSRP.

Fake msrp

GPU MSRP seems to have no value these days. There are no MSRP RTX 50 models currently in the stock, and it seems that it will not appear on the shelves in the future.

If you go to ASUS online storeYou will see that you can not even find the RTX 5080 at $ 1,000 from MSRP. The most affordable option, RTX 5080 Prime, Prime is the ASUS “MSRP” series, costs exactly $ 1,264.99 at the time of writing this report. Msi Don’t even sell MSRPOW RTX 5080 anymore. The only shade model you can find is the OC Pricier model that sells it for $ 1,249.99.

The position with RTX 5090 cards is worse, as the most expensive forms cost hundreds of dollars at the demand price of $ 2000 for the card (if you can find one).

RTX 5070 TI and RTX 5070 can be found sometimes in MSRP if you are fast enough to order one before you get robots and provinces, or if you are lucky to have a small center near you. But given what happened to the MSRP RTX 5090 and RTX 5080 cards, I will not be surprised if we soon see MSRP RTX 5070 TI and RTX 5070 disappear forever.

Finally, it would be better if I hadn’t even talking about the situation in Europe, Australia and other markets, where MSRPS is higher, and there are no MSRP cards anywhere.

Medium -range graphics processing budget and 8 GB memory and 12 GB of memory

While NVIDIA has equipped the RTX 5070 TI with 16 GB memory, the RTX 5070 stuck with a 12 GB temporary memory store becomes very slow for games at 1440 pixels and high details. Run Ray Tracking in modern AAA games, and will approximately use of memory. RTX 5070 seems to have a similar fate for the RTX 3070, which was 8 GB memory very few for games with high or super pictures shortly after the card appeared.

The position with the RTX 5060 looks more clear. Kopite7kimi, a well -known leak with a flawless busy record, I participated recently RTX 5060 specifications. It seems that the card will only get a memory of 8 GB, which is unacceptable in 2025 unless we talk about the graphics processing unit below $ 200, which will definitely not be the RTX 5060. Remember that 8 GB of memory first appeared on the GTX 1070, which was launched about nine years ago!

My direct experience is that the memory 8 GB is not enough even for games 1080 pixels with the quality of the maximum texture, even without tracking the rays. I had to drop the textures to the average The horizon is forbidden to the West When the game streams on ASUS ROG ally to avoid a decrease in the huge performance as soon as the temporary store of memory is filled about 20 minutes after playing the game with a texture. I had similar experiences in 2 Heard and Forza Horizon 5So if the RTX 5060 appears for the first time with 8 GB memory, it will be doa.

On the other hand, inconvenient The retail menu leakage It claims that the card will come with 12 GB memory. I hope Kopite7kimi is wrong and that RTX 5060 will already pack 12 GB memory. Otherwise, anyone who ends up will be in a bad surprise.

Laced units of embodiment and melting energy cables

The last nails in the RTX 50 coffin are the lost ROPS (output units) and the height of the melting power cable.

Some RTX 5090, RTX 5080 and RTX 5070 TI cards are flawed in that they lack a certain number of ROPS, making the affected cards up to 10 % in games, according to Nexus players.

This is a dangerous supervision by Nafidia. The company somehow failed to notice the problem during the quality guarantee process. The number of affected cards is small, but this does not reduce the risk of this failure. To add an insult to the infection, while NVIDIA claims that the RTX 50 graphics processing unit is not affected by Jarrodtech (one of the best laptops there) claim NVIDIA asked them to check their own laptops.

Fortunately, the RTX 50 Mobile graphics processing units do not suffer from the lost ROPS. However, the fact that NVIDIA asked manufacturers to verify laptop tools that do not paint a beautiful image in relation to the quality guarantee process.

The second stirring includes soluble energy cable Very bad uprightBut the 12V-2×6 connector that replaced it It is worse.

Power cables will continue to melt unless NVIDIA will replace 12V-2×6 conductor or redesign RTX 50 circuit panels to include a better wutter protection solution. Currently, everything indicates that NVIDIA will not do it.

The only silver lining may be that the RTX 5090 cards have not been found for weeks now. There is an opportunity for the company to try to solve the problem under the radar, and that the future RTX 5090 cards will not come with poor energy connectors.

For me, I got AMD RX 9070 XT instead of RTX 5070 TI. The AMD card provides a greater value with an excellent solution in FSR 4, which finally matches DLSS in image quality.

If NVIDIA offers more uproar to Back compared to AMD with the next generation of gaming graphics processing units (I doubt this), I will replace because I do not have any brand. But the fact of the matter is that the RTX 50 series is a tremendous disappointment, and it is the largest of NVIDIA yet, as much as I am concerned.