Nvidia’s AI Chips Are All the Rage. However, Gamers Are Still Complaining About Issues With Its Graphics Cards Months Later

In 1993, three engineers in Silicon Valley came together with an idea that would eventually transform the video game industry. Current Nvidia CEO Jensen Huang, Chris Malachowsky, and Curtis Priem aimed to create a chip to enhance PC graphics. From that meeting, Nvidia was born, along with the GPU as we know it today.

For years, Nvidia has been synonymous with the PC gaming experience. If you’ve ever built a gaming computer, chances are your graphics card features its logo. The brand has achieved something even more challenging than technological innovation: cultivating a loyal community that eagerly anticipates each new release.

No Longer What It Used to Be

However, since the introduction of the RTX 50 series, with models such as RTX 5090 and RTX 5080, user complaints have increased. Many gamers have encountered bugs in the drivers, and the company has been slow at providing them with solutions. While it’s sometimes possible to revert to an earlier driver version to avoid these issues, users with the latest graphics cards can’t. The RTX 50 models are incompatible with the more stable previous versions, like 566.36.

Reddit has become a hub for these complaints, with threads filled with testimonials from frustrated users. YouTube has also amplified the issue, featuring videos that document these problems. This is more than just chatter on social media. Nvidia has acknowledged dozens of errors in its latest drivers and has promised to address them in upcoming updates.

In addition to the officially recognized issues, complaints about stability problems have also been common among users of previous generations. Several gamers have reported black screens, unexpected reboots, flickering, visual artifacts, and loading errors. Others have complained about forced shutdowns in games such as Cyberpunk 2077 and Star Wars Outlaws.

Nvidia’s headquarters in Santa Clara, California.

More from Xataka On

Nvidia Will Continue to Sell Its H20 AI GPUs in China. It Cost the Company a $1 Million Dinner

Some of these failures are tied to specific configurations, such as when using monitors with high refresh rates or having G-Sync enabled. Others are linked to using particular DisplayPort outputs, which complicates diagnosis.

This comes after a rocky debut. Some power cables on the RTX 5090 have melted down. Shortly after the launch, some users also discovered that their RTX 50 series graphics cards had fewer ROP units than the official specs indicated. Nvidia has confirmed that this manufacturing defect affects less than 0.5% of chips in models such as the RTX 5090, 5090D, 5080, and 5070 Ti. According to the company, the average performance impact is 4% and doesn’t affect AI or computational tasks.

Meanwhile, the role of graphics in Nvidia’s ecosystem is evolving. While they remain important, the company is now focusing primarily on artificial intelligence. This shift is reflected in its press releases and the design of its website, which has changed significantly since 2012 and even 2018.

This isn’t the first time a major tech company reconfigures itself. Microsoft had its core in Windows for years, but has since diversified its product portfolio and business units. This decision has been crucial in helping the company remain competitive over time. The same is true for Nvidia. Founded in 1993, Nvidia continues to grow and adapt. More importantly, part of its strength lies in its strategic choices. Despite mistakes and successes, it’s always remained active and evolving.

Images | Đào Hiếu | Coolcaesar (Wikimedia Commons)

Related | Nvidia Eyes Another Disruptive Technology After the Success of Its AI Chips: Quantum Computers