Intel releases XeSS 2.0 developer tools, but keeps code under lock and key

By Kunal Khullar Published March 19, 2025

Three main components of Intel's XeSS 2 upscaling — Intel

Intel has finally released the XeSS 2.0 Software Development Kit (SDK) to aid game development through upscaling, frame generation, and latency reduction technologies. The SDK is now accessible via Intel’s GitHub repository, providing developers with pre-built tools for integrating XeSS into their projects.

Intel expanded its suite of XeSS technologies last year when it unveiled its budget segment GPUs, the Arc B580 and B570. The XeSS 2.0 SDK introduces three main components: XeSS-SR for AI-driven super resolution, XeSS-FG for frame generation, and XeLL for low-latency improvements. These advancements position XeSS 2.0 as a competitor to Nvidia’s DLSS (Deep Learning Super Sampling) and AMD’s FSR (FidelityFX Super Resolution), offering similar AI-powered enhancements to image quality and performance. Intel has also updated the XeSS inspector tool to support Vulkan and DirectX 11, making it easier for developers to optimize game performance across multiple rendering APIs.

Recommended Videos

To help with integration, Intel has developed plugins for widely used game engines, including Unreal Engine and Unity. However, despite these efforts to make the SDK more accessible, Intel’s decision to keep XeSS 2.0 closed-source limits its flexibility compared to AMD’s open-source FSR.

As spotted by Tom’s Hardware, the licensing agreement prohibits reverse engineering, decompilation, or disassembly, preventing developers from modifying or extending its capabilities beyond what Intel provides. This approach means that developers must work with pre-compiled binaries, which could impact broader adoption, particularly among studios seeking greater control over their graphics pipelines.

Additionally, XeSS 2.0 remains a Windows-first technology, restricting its immediate availability on other platforms. While Intel has not ruled out future cross-platform support, its current design prioritizes Windows-based gaming environments. The technology has yet to see widespread adoption, with only a handful of titles currently supporting XeSS 2.0. Intel has acknowledged this limitation and is actively working with game developers to expand the presence of XeSS in upcoming releases.

Former Computing Writer

Kunal Khullar is a computing writer at Digital Trends who contributes to various topics, including CPUs, GPUs, monitors, and…

Topics

Tech News

Computing

I didn’t think fake shopping could trick my brain until I tried the viral dopamine websites

On these fake shopping and delivery sites, nothing ever ships or costs a cent, yet the dopamine hit still felt real.

fake-shopping-on-viral-dopamine-websites

I spent some time in the internet's fakest mall, filling a shopping cart with things I could not buy, hunting for discounts on products that do not exist, and checking out three separate times for a grand total of $0.00. Then I placed a fake food order and smoked a cigarette I could not taste, with strangers I will never meet, on a rooftop that does not exist either. My bank balance never changed, but my brain, annoyingly, did.

When I first heard about South Korea's growing "dopamine sites," I assumed they were another internet oddity I would poke around for five minutes and forget by lunch. Instead, I found a surprisingly clever idea hiding beneath the absurdity. You can browse endless imaginary products, add everything to your cart, and complete a purchase that never actually exists. These websites aren't trying to sell you anything. They're trying to recreate the feeling of shopping while removing the part that empties your wallet.

Computing

LG wants to build humanoid robots, and NVIDIA is giving it the brains

This isn't just another robot teaser. LG and NVIDIA just outlined actual hardware.

Robot, Appliance, Device

With time, more and more legacy hardware companies seem to be racing to bolt a humanoid robot onto their roadmap. The latest entrant is LG, joining the race with tentative timelines.

The company's bipedal humanoid, built on Nvidia's robotics stack, is planned for early 2027, and it's backed by a broader push into AI factories and self-driving vehicle platforms.

Computing

Googlebook may use older Snapdragon chips to build more affordable laptops

Snapdragon X Plus and X Elite models could be in development

Googlebook

Google has already confirmed that Googlebook will support more than one chipmaker, and a new leak may have revealed the first Snapdragon models in development.

New code references uncovered by GbookHub reportedly link two Googlebook designs, codenamed Annite and Pic, to Qualcomm’s Snapdragon X Plus X1P-42-100 processor. Both are also said to be tied to a board platform called Mica.