Machines with RTX are artificial intelligence computers; We tested a laptop with 4060 [Review]

Since 2022, AI has become synonymous with the latest and most innovative technology to reach the domestic market. To increase sales and bring an air of innovation, the industry AI PCs for the stage, that is, a set of computers with special chips to perform artificial intelligence-related tasks.

But the truth is Artificial intelligence computers have been around for a while, and Nvidia is one of the ones responsible for this.. To better understand this story TecMundo By testing the Avell laptop equipped with GeForce RTX 4060, which focuses on artificial intelligence activities, it proved that the green team is ahead in this regard.

RTX revolution

Before we get into the user experience, it’s worth going back in time and explaining how GeForce RTX not only revolutionized Nvidia’s product portfolio, but also pioneered a number of innovations in the field of artificial intelligence.

Released in September 2018, RTX 20 came with a number of features previously unknown to the public. The iconic GTX (Giga Texel Shader eXtreme) prefix, which has been used for years, has been replaced by RTX (Ray Tracing Texel eXtreme), which will stand out in the industry.

With this generation, the company introduced Tensor Cores, or Tensor Cores in Portuguese. These cores are specialized structures that enable hybrid-aware computing to enhance productivity scenarios, high-end gaming, and even deeper execution of AI-related activities through machine learning.

Thus, Nvidia was able to work on technologies that could use artificial intelligence and LLMs (large language models) on a large scale in the coming years. A clear example of these resources is DLSS (Deep Learning Super Sampling), which uses artificial intelligence and machine learning to upscale the image to deliver higher performance in games.

The biggest feature of tensor cores and other technical structures found in all RTXs is to facilitate the execution of these applications directly on the graphics card hardware. If it was previously necessary to resort to the cloud to use artificial intelligence tools or much more complex components, The arrival of RTX has caused native GPUs to enter the fray.

Do RTX and AI work well?

I had the opportunity to test the Avell A65 Ion with the RTX 4060 as my main laptop for about a month and perform many practical tests with the AI. Although the laptop is not registered as a Copilot+ PC, it can basically handle any workload related to this subject without any problems and very quickly as expected.

To start the test batch, I used the latest version of Photoshop, packed with Adobe’s AI resources. Unlike some AI computers I’ve tested in recent months, the Avell model can perform Fill and Productive Expansion functions faster and without crashing.

Thus, it was possible to correct the photographs I took for analysis. TecMundoRemoving defects in images such as exposed wires or shadow problems caused by lighting. Since we don’t have an “official” AI PC to test on, graphics, speed, etc. It is not possible to make a direct comparison with , but the idea lies in the user experience.

I had the opportunity to test the software’s native image production with the command “A car and helicopter hybrid based on Tesla and Porsche models” in Photoshop. The app created realistic images in less than ten seconds and without causing the machine to drownas it usually is.

After the synthetic images, I wanted to test Adobe Firefly to see how the tool would behave on this laptop. When the command “A cyberpunk-style mechanical shark with red eyes and tattoos” was applied, the result was equally satisfying.

Nvidia Canva

One Nvidia technology I’ve never been able to use is Nvidia Canva, which turns simple drawings into a Paint-like space to create realistic landscapes and terrains. When I had already installed the software, I decided to create a beach with some type of dry vegetation, certain plants, a strip of sand and of course the sea.

The combination of these elements is very interesting, but it is not 100% perfect as a command implemented in Copilot or Photoshop itself. The most interesting thing is the smoothness with which image components are rendered almost instantly. In every line you write, The AI ​​built into Canva immediately turned it into something relatively realistic.

Chat with RTX

If OpenAI has ChatGPT and Google has Gemini, Nvidia has found a way to build its own chatbot through ChatRTX. This is another green team technology I’m testing for the first time and I have to say I’m disappointed.

The interface is similar to traditional assistants, but the software only works in English. ChatRTX, “How to choose the best graphics card to run games in 2025?” He even understood my command. It’s in Portuguese, but I sent the response in the native language.

The biggest problem is the delay in responses. chatbot lasted much longer than competitorslike native Windows Copilot. It’s hard to tell whether this is a problem with the technology itself or whether Nvidia’s AI just doesn’t work very well with chatbots, which is highly unlikely.

What about games?

As explained in the first title of this article, Nvidia cards have been compatible with DLSS technology since 2018. Although not widely considered an AI technology, DLSS is equipped with AI to make games have better graphical quality by reducing aliasing and presenting more frames on the screen.

as it is TecMundo We are very curious about the performance, of course we would run some tests in games to understand the capabilities of DLSS. In this benchmark, I ran Cyberpunk 2077 at maximum quality, Quad HD+ resolution (2,560×1,600) and Ray Tracing on ultra with Nvidia’s upscaling technology.

The result is a highly visible difference when using the company’s AI technologies. By simply enabling DLSS in balanced mode, players can achieve a much more stable frame rate range close to 40 FPS. The situation is even better when Frame Rendering is enabled, which uses motion vectors to create new frames. improve performance by almost 60%.

RTX are artificial intelligence computers

RTX graphics cards developed to run games and professional applications show that they can be included in the AI ​​PC concept. The major difference in this terminology is that AI PCs are currently called that because they have a built-in NPU (Neural Processing Unit). The processor chip and the GPU already have these structures on their own chip.

As a theoretical comparison, AMD’s Ryzen AI 300 processors have a maximum of 50 TOPS (Trillion Operations Per Second) on the processor alone, regardless of the capabilities of the dedicated or integrated graphics card in the laptop. The RTX 4060 laptop we tested alone has 233 TOPS.

As with all new technologies available on the market, with absurdly expensive prices at the beginning of the cycle, Computers with artificial intelligence still have a big place in the electronics market. However, Nvidia graphics cards for desktop or laptop computers are widely available for different audiences at prices ranging from R$ 1,800 to almost R$ 30,000 in Brazilian retail.

Like it or not, AI PCs are here to stay, but there’s no denying that Every computer with GeForce RTX is also an AI machine. Whether it’s games, studies, or professional applications, the industry has been working with AI for a while now, and you may already have this technology at home.