5 worst models

5 worst models

Nvidia is a very successful corporation, one of the richest in the world. For its products, especially chips like H100 and A100, queues of representatives of Google, Microsoft, etc. line up. All this indicates that Nvidia’s management understands how to work correctly, engineers are competent, and excellent salespeople. But even such a company has failures. For example, video cards. Today we will talk about the worst graphics adapters from Nvidia in its history.

GeForce RTX 3080

At first glance, it’s a good card. For the RTX 30 series, Nvidia developed a new Ampere architecture that helped solve many of the problems of the Turing-based RTX 20 series chips. In addition, chips began to be produced on the 8-nm process, which meant greater performance with improved energy efficiency.

In September 2020, the company launched the RTX 3080 and 3090 models. The $699 3080 was roughly on par with the $1,499 3090. In addition, this card outperforms the RTX 2080 Ti in both ray-traced and non-ray-traced games . Many experts rated the map positively because it was really good.

Why is she on the list? After the release of the video card, it turned out that it is almost impossible to buy it at the recommended price. The coronavirus pandemic and the cryptomining epidemic have led to a sharp increase in the price of cards. The price of the model rose above $1,500. Even after the end of the pandemic, two years after the release of the card, its price did not drop below $699.

When ordinary gamers were finally able to test it, that is, after the end of the deficit, it turned out that the adapter lacks memory, especially for the latest titles. In addition to everything, a number of cards of this series are equipped with budget cooling, which is clearly not enough for a powerful chip.

In general, all this made the GeForce RTX 3080 one of the most unsuccessful in the company’s history. The vast majority of gamers have never met her.

GeForce RTX 4060 Ti

Despite the fact that this is a relatively new product, the video card went on sale only six months ago, it can already be called a failure. At the same time, it has positive qualities. For example, the video adapter is more productive than the 30 series cards. Also, the chip layout is three times denser than that of the GA104. But there are also problems, and most of them are again related to memory.

There are two variants of the 4060 Ti model: one comes with 8GB of video memory, the other with 16GB. Obviously, 8 GB is clearly not enough anymore, especially since we just talked about similar problems with the previous generation of adapters. But the GeForce RTX 4060 Ti with 16 GB costs $100 more, which already hits your pocket hard.

But there are also savings on small things. For example, a 128-bit bus is installed, which means half the bandwidth of the previous generation cards – 288 GB/s instead of 448 GB/s. Also, the 4060 Ti only has eight PCIe lanes instead of 16.

As a result, this model is an analogue of the RTX 3070 without increased performance and support for DLSS 3 Frame Generation. Thus, we got an expensive analogue of the previous generation of cards without much improvement. Everything would be fine if Nvidia released these models with a video memory capacity of 12 GB and a 192-bit bus. But no, the manufacturer decided to save money, as a result, the card was unsuccessful.

GeForce RTX 2080

Adapters of the 20th RTX series could be called a breakthrough in gaming. The fact is that these are the first video cards with ray tracing. This technology even began to be called the “Holy Grail” of graphics in games.

But for this revolution in design, it was necessary to pay, and not to the manufacturer, but to the buyers. Yes, RTX series cards were very expensive. For example, the RTX 2080 Ti debuted at a price of $1199. None of the company’s video cards had previously been sold at such a high price.

In addition, at the time of the release of the graphics adapter, there were practically no games with support for ray tracing. And a lot of time passed before such tiles appeared. The only game of 2018 to offer ray tracing, Battlefield V, didn’t show off the full benefits of the technology.

In general, the new graphics adapters were more productive than the previous generation video cards. But they cost so much that many gamers preferred not to upgrade, there was no direct benefit from the upgrade, but it had to be spent quite a lot.

By the way, even many months after the release of this generation of adapters, ray tracing has only been added to a small number of games. The same Atomic Heart, despite numerous demos with support for the technology, came out without a trace, the developers promised to add all this later. Statistically, even today only 37% of RTX 20 series users have ray tracing enabled.

What was really groundbreaking in the 20th series was Deep Learning Super Sampling, or DLSS. The start of the technology was also difficult, but today it is integrated into more than 500 games and works perfectly. Maybe Nvidia should pay more attention to DLSS than ray tracing.

In general, the main disadvantage of this generation of graphics adapters is an unreasonably high price. If it weren’t for this, the card would hardly make the list.

GeForce GTX 480

And these are matters of the past days. And at that time it was a flagship, but very hot, in the literal sense of the word. The energy efficiency of the chip was extremely low, the card consumed about 300 W, which corresponded to the level of energy consumption of the competitor HD 5970 with two graphics processors at once.

The outline of the standard cooler somewhat resembled a grill, which became a reason for mockery and mimes. Videos were created with cooking using a video card, etc. In general, this was not a good time to promote these adapters. Overheating affected the failure of cards, which was very frustrating for gamers.

In the end, the company had to urgently give other cards to the market, they became GTX 500 series adapters with a modified Fermi architecture. One of them, the GTX 580, also consumed 300 W with a performance increase of only 10%. However, the rest of the models of the new series had better energy efficiency.

The setback affected the company’s market, with AMD then coming very close to overtaking Nvidia, something it hadn’t done since the mid-2000s. Despite all this, Nvidia still remained a very profitable company, so it managed to get out of the water almost dry. Moreover, using the experience of failure, it was able to improve the quality of its products in subsequent generations of adapters.

GeForce FX 5800 Ultra

At the beginning of 2003, Nvidia had problems – the AMD company released excellent graphics adapters of the Radeon 9000 series on the market. The chip was produced using a 150-nm process, Nvidia managed to release a worthy competitor only after switching to a 130-nm process.

But in an effort to introduce an alternative as soon as possible, as well as to make its production cheaper, the company left the card with a low bandwidth of 16 GB/s compared to almost 20 GB/s in the 9700 Pro. There was also an issue where the anisotropic filtering didn’t look very good, which was another drawback of the new adapter.

But the main disadvantage is that the card had a huge for that time power consumption of 75 W. The installed fan was so noisy that this sound was compared to the noise of a vacuum cleaner or a leaf blower. Then memes appeared on this topic – one of them is posted above.

The shortcomings led to the fact that Nvidia quickly began the development of a new adapter. And this time it all worked out — a new, improved GeForce FX 5900 Ultra flagship went on sale six months later. The GPU had a massive 27GB/s bandwidth for the time, patched drivers and much more reasonable power consumption of around 60W.

And which video card from Nvidia from those models that you tested yourself would you call unsuccessful?

Related posts