Skip to content
Tech News
clear
Topics: Today This Week This Month This Year

Valve's Steam Frame VR headset launches focused purely on gaming

Valve has released the Steam Frame, a standalone VR headset built specifically for playing games rather than serving as a multipurpose productivity or mixed-reality device. Early reviews suggest the device leans into a lighter, more playful approach to virtual reality rather than chasing the enterprise and work-focused ambitions of rivals like Apple's Vision Pro.

Valve opens Steam Frame reservations at $1,049 for its return to standalone VR

Valve has begun taking reservations for the Steam Frame, a standalone VR headset priced at $1,049, marking its first major hardware push into virtual reality since largely stepping back from the category years ago. Ars Technica's early hands-on testing found the device notably lightweight at 440 grams, about 15 percent lighter than the Meta Quest 3, thanks to a split design separating the battery pack from the lens housing.

Valve's Steam Frame headset launches with standalone VR and PC streaming for over $1,000

Valve has released the Steam Frame, a VR headset that runs on a Qualcomm Snapdragon 8 Gen 3 chip and offers up to 1TB of storage, letting it play games independently without a connected PC. The device is significantly lighter than the older Valve Index, weighing 440 grams versus 809 grams, and uses six cameras for tracking, passthrough, and eye-tracking. It arrives as the third major hardware release from Valve in 2026, following the Steam Controller and Steam Machine.

Valve's Steam Frame VR headset ships with standalone Snapdragon power for $1,000-plus

Valve has released the Steam Frame, a VR headset that can run games independently using an onboard Qualcomm Snapdragon 8 Gen 3 chip and up to 1TB of storage, rather than relying solely on PC streaming like its predecessor, the Valve Index. The device weighs roughly half as much as the Index thanks to lighter materials, and uses six cameras for tracking, passthrough, and eye-tracking, with a price tag exceeding $1,000.

Valve prices Steam Frame VR headset from $1,059, bundles rebuilt Half-Life: Alyx

Valve has revealed pricing for its Steam Frame VR headset: $1,059 for the 256GB model and $1,299 for the 1TB version, both including controllers and a version of Half-Life: Alyx rebuilt to run natively on the device. Optional accessories such as a charger, replacement gaskets, ergonomic straps, prescription lenses and an add-on RGB camera will also be sold separately.

Valve's Steam Frame headset runs Half-Life Alyx standalone, but review calls it late to market

Engadget's review of Valve's Steam Frame VR headset finds it capably runs a standalone version of Half-Life: Alyx along with older VR titles like Beat Saber and Superhot, loading faster than Meta's Quest 3. The headset also supports newer SteamVR games, 2D Steam titles, and wireless PC streaming, positioning it as a broader 'Steam Machine' for the face rather than a pure standalone device.

Binomial LLC Explores Evolution Strategies for Fitting Neural and PBR Textures Without Backpropagation

A Binomial LLC engineer working on GPU texture interchange has published research fitting neural textures and physically based rendering material maps using evolution strategies (ES) instead of traditional backpropagation. The approach avoids gradient-based optimization typically used in neural texture compression and material fitting workflows.

Study probes why AI research agents resist overfitting despite iterative benchmark tuning

The piece examines a paradox in machine learning practice: researchers repeatedly check performance on held-out validation data while iterating on model design, a process that in theory should contaminate that data and cause overfitting. Yet in practice, research agents and human researchers alike often generalize well despite this repeated peeking, prompting a deeper look at why the standard overfitting warning doesn't always play out as expected.

Explainer breaks down how truncated SVD compresses image data using PCA math

A technical write-up walks through the relationship between Principal Component Analysis and Singular Value Decomposition, showing how a matrix can be factored into three components (U, Sigma, V) to capture the most significant variation in data. It uses a 1024x1024 grayscale image of the moon as a worked example, treating the image as a set of vectors that can be decomposed and reduced via truncated SVD.

Today's top topics: openai anthropic apple ai safety iphone 18 pro dario amodei ios 27 artificial intelligence google nvidia
View all today's topics →