Nvidia DLSS 4.5 AI Performance Takes Center Stage as Hardware Skips CES Keynote
Nvidia focused on AI and DLSS 4.5 performance at CES 2026, skipping new GeForce hardware. Learn how the new transformer model improves upscaling quality.
The GPUs are missing from the stage, but the AI isn't. For the first time in years, Nvidia declined to introduce new GeForce graphics card models during its CES keynote. CEO Jensen Huang delivered a 90-minute presentation focused almost exclusively on the company's enterprise AI dominance, leaving gaming announcements for a separate, secondary digital release.
Analyzing Nvidia DLSS 4.5 AI Performance Gains
Instead of raw silicon, the focus shifted to software-driven optimization. The headline act for gamers is DLSS 4.5, the latest iteration of Nvidia's deep learning super sampling. According to company technical leads, this version introduces a 'second-generation transformer model' trained on a significantly expanded dataset.
Nvidia's Bryan Catanzaro noted that DLSS 4.5 is specifically designed to tackle image quality issues in aggressive upscaling modes. By improving pixel prediction in Performance and Ultra Performance settings, the technology allows gamers to squeeze more out of their existing RTX hardware without needing an immediate upgrade.
The Strategic Shift to AI-First Gaming
This pivot reflects a broader industry trend where software becomes the primary driver of generational performance leaps. While enthusiasts may be disappointed by the lack of a 'GeForce 50-series' or similar reveal on the main stage, the focus on DLSS 4.5 highlights how AI is now the bedrock of Nvidia's consumer strategy.
Authors
Related Articles
Nvidia shipped roughly a billion RISC-V cores in 2024, then announced it would run CUDA on the open standard. We break down how royalty-free instruction sets and open software stacks are trying to route around CUDA's lock-in. Part 2 of the Semiconductor Sovereignty series.
US AI-chip export controls split into three layers in the first half of 2026 — January easing, a May crackdown on circumvention, and a pending bill. Nvidia erased China from its guidance and still posted a record $81.6 billion quarter. A look at the export policy that both shields and cages it.
AMD's MI325X matches or beats Nvidia on memory and bandwidth — yet Nvidia's 86-92% share holds. The real moat is CUDA, 20 years in the making. Part 1 of 4.
Snowflake's new $6 billion AWS contract is about more than cloud spending. It signals a shift in AI infrastructure—away from Nvidia GPUs and toward cheaper, homegrown chips for the agent era.
Nvidia shipped roughly a billion RISC-V cores in 2024, then announced it would run CUDA on the open standard. We break down how royalty-free instruction sets and open software stacks are trying to route around CUDA's lock-in. Part 2 of the Semiconductor Sovereignty series.
US AI-chip export controls split into three layers in the first half of 2026 — January easing, a May crackdown on circumvention, and a pending bill. Nvidia erased China from its guidance and still posted a record $81.6 billion quarter. A look at the export policy that both shields and cages it.
AMD's MI325X matches or beats Nvidia on memory and bandwidth — yet Nvidia's 86-92% share holds. The real moat is CUDA, 20 years in the making. Part 1 of 4.
Snowflake's new $6 billion AWS contract is about more than cloud spending. It signals a shift in AI infrastructure—away from Nvidia GPUs and toward cheaper, homegrown chips for the agent era.
Thoughts
Share your thoughts on this article
Sign in to join the conversation