AMD launches FSR Redstone with AI frame generation to directly challenge Nvidia DLSS dominance

Reviewed byNidhi Govil

8 Sources

Share

AMD released its FSR Redstone driver update, introducing machine learning-powered frame generation and ray tracing features for RDNA 4 graphics cards. The update brings AMD closer to feature parity with Nvidia DLSS 4, though it arrives seven months after its initial announcement and currently supports only around 30 games with full AI capabilities.

AMD Shifts to Machine Learning with FSR Redstone

AMD has officially launched FSR Redstone, a comprehensive AMD driver update that marks a fundamental shift in the company's approach to gaming upscaling technology. After months of anticipation since its Computex 2025 announcement, the update introduces machine learning-powered features that position AMD as a serious contender in competing with Nvidia and its established DLSS 4 technology

1

3

. The Redstone update represents AMD's transition from analytical-based algorithms to neural network-driven solutions, implementing AI frame generation, neural radiance caching, and FSR Ray Regeneration technologies that leverage the specialized hardware in RDNA 4 graphics cards

1

.

Source: IGN

Source: IGN

The shift to machine learning addresses longstanding criticisms that AMD's previous FSR versions lagged behind Nvidia's offerings in improved image quality and performance. By embracing AI-driven rendering, AMD aims to eliminate the ghosting and artifacts that plagued earlier iterations while delivering enhanced ray tracing performance through more efficient processing

3

.

Understanding the Core Technologies in FSR Redstone

FSR Redstone bundles four distinct features under a unified platform, though AMD's naming conventions have created confusion among PC gamers. FSR Upscaling, formerly known as FSR 4, now uses machine learning instead of analytical methods to scale lower-resolution frames to higher resolutions while maintaining visual fidelity

4

. The new AI frame generation technology represents the most significant advancement, using neural networks to insert generated frames between real frames, dramatically improving frame rates. Early testing shows this approach virtually eliminates the weird artifacts present in FSR 3's analytical frame generation

5

.

FSR Ray Regeneration works to remove noise and enhance ray tracing details at lower rendering costs, functioning as AMD's answer to Nvidia's ray reconstruction technology

1

. Meanwhile, Radiance Caching enables GPUs to maintain a small cache of rays for AI models instead of tracing each ray for every pixel, significantly reducing render time in path-traced games

1

. However, Radiance Caching won't appear in games until 2026, making it the only Redstone feature not yet available

4

.

Performance Gains and Hardware Requirements

AMD claims substantial performance improvements with FSR Redstone enabled. In Call of Duty: Black Ops 7, the company demonstrated frame rates jumping from 23 fps to 109 fps when using performance settings with resolution scaling and frame generation

2

. However, AMD recommends games already run at a minimum of 60 fps after upscaling to avoid artifacts, meaning these technologies primarily enhance already playable experiences rather than salvaging poor performance

1

2

.

Source: Tom's Guide

Source: Tom's Guide

The catch is that all machine learning features in FSR Redstone require RDNA 4 graphics cards, specifically the Radeon RX 9000 series including the RX 9070 and RX 9070 XT

2

3

. Older RDNA 1 through 3.5 GPUs remain limited to analytical versions of upscaling and frame generation, missing out on the neural network advantages

1

. This hardware restriction puts AMD at a disadvantage compared to Nvidia, which supports DLSS 4 features on cards dating back to the RTX 30-series

3

.

Game Support Remains a Critical Challenge

While AMD promises over 200 games will support FSR Redstone by the end of 2025, the reality is more nuanced

2

3

. Only around 30 to 40 games currently support the crucial machine learning-powered frame generation feature, with titles like Black Myth: Wukong, Cyberpunk 2077, and F1 24 among the early adopters

2

5

. Radiance Caching and FSR Ray Regeneration have even more limited availability, with only Warhammer 40K: Darktide and Call of Duty: Black Ops 7 supporting these features respectively

2

.

Source: PC Gamer

Source: PC Gamer

This represents a massive gap compared to Nvidia DLSS, which boasts compatibility with over 800 games

3

. The disparity highlights AMD's challenge in convincing developers to implement its technologies, though the company promises more games will arrive in 2026

2

. For now, users must manually enable Redstone features through AMD's Adrenalin Software on a per-game basis, adding complexity to the user experience

5

.

Closing the Gap with Nvidia but Missing Key Features

FSR Redstone brings AMD closer to feature parity with Nvidia DLSS 4, but notable differences remain. AMD's approach uses a 1:1 ratio for frame generation, inserting one AI-generated frame between each rendered frame

3

. In contrast, Nvidia's multi-frame generation technology offers a 3:1 ratio on RTX 50-series cards, potentially delivering smoother performance

1

3

. The absence of multi-frame generation in Redstone means AMD hasn't fully matched its competitor's capabilities.

However, AMD maintains one advantage: its modular approach allows users to selectively enable individual features rather than forcing adoption of the entire suite, as Nvidia does

1

. Early testing suggests the quality of AMD's AI frame generation delivers substantial improvements over FSR 3, with testers noting that artifacts and blurring during rapid camera movements have been virtually eliminated

5

.

Implications for Console Gaming and Future Hardware

The significance of FSR Redstone extends beyond PC gaming. Sony collaborated with AMD on developing FSR 4, which launched in March, and the company has indicated plans to bring the technology to PS5 Pro

2

. While the original PS5's RDNA 2 architecture presents challenges for implementation, speculation suggests Redstone could prove transformative for the rumored PlayStation 6, which would likely incorporate RDNA 4-based internals

2

. Upscaling technology benefits lower-end systems more dramatically than high-end hardware, meaning future gaming handhelds from Sony or PC manufacturers could see substantial performance gains from these advancements

2

.

The potential integration with SteamOS also presents intriguing possibilities. With AMD-powered machines running Valve's Linux-based operating system, the combination of FSR Redstone and SteamOS could create compelling alternatives to traditional Windows-based gaming PCs

3

. This matters because it could accelerate adoption of AMD's GPU technologies across multiple gaming platforms, from handhelds to consoles to PCs.

What This Means for the GPU Market

FSR Redstone arrives seven months after its initial announcement, giving Nvidia DLSS 4 nearly a year of market dominance

5

. The delayed launch and limited initial game support suggest AMD faces an uphill battle, yet the technical quality of the implementation demonstrates the company's commitment to catching up. For consumers, this competition drives innovation in gaming performance technologies, potentially making high-quality gaming experiences more accessible across different price points. The focus on machine learning and neural networks signals that AI will continue shaping how games render and perform, with implications for hardware requirements and game development practices going forward. AMD's success with Ryzen processors against Intel shows the company can execute comeback strategies, and FSR Redstone represents a critical step in applying that same approach to the GPU market

1

.

Today's Top Stories

TheOutpost.ai

Your Daily Dose of Curated AI News

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Spend less time searching for the latest in AI and get straight to action.

© 2026 Triveous Technologies Private Limited
Instagram logo
LinkedIn logo