SAN JOSE, California – Nvidia CEO Jensen Huang came out in defense of the company’s DLSS 5 AI model for boosting gaming graphics, saying those criticizing the offering were “completely wrong.”

During a press conference attended by SDxCentral, the CEO was bullish, saying, “we created the technology. We don't create the art.”

DLSS 5 was the first thing unveiled during Hanug’s two-hour-long keynote and the only gaming-related offering in a showcase dominated by AI infrastructure capabilities, including the new Groq 3 language processing unit (LPU) and its own OpenClaw offering, dubbed NemoClaw.

But the sole gaming reveal was quickly lampooned on social media, with a ton of memes and some users decrying it as “AI slop.”

Huang defended the offering, saying it was “very different than generative AI.”

“DLSS 5 is fine because it's geometry control. It's conditioned by the ground truth of the game. It enhances and adds generative capability to it, but it doesn't change artistic control," Huang said. “It's not post-processing. It's not post-processing at the frame level. It's generative control at the geometry level. Everything about the generative part to demonstrate the technology and see what you do if you decide you want it to be a cartoon tune shader, or made of glass. It's in the control, direct control of the game.”

Another vocal proponent of the technology was Todd Howard, executive producer at Fallout and Skyrim makers Bethesda Game Studios. Howard said the technology, which was demoed at GTC on the studio’s Starfield game, was “amazing.”

A demo running on the GTC show floor, seen by SDxCentral, required two GPUs – one for the graphics and one just to power the model. Nvidia staff on site were keen to stress it was just a demo and that once fully released will not require two 50-series graphics cards.