AMD’s “FSR Redstone” upscaling claims to close the gap with Nvidia’s DLSS

AMD unveiled FSR Redstone, a suite of four ray-tracing and frame-generation technologies (three new, one renamed) designed to raise FidelityFX Super Resolution image quality closer to Nvidia’s DLSS while remaining straightforward for developers already using FSR 3.1 or 4. Unlike prior FSR releases that ran on a wide range of GPUs, Redstone will be limited to AMD’s latest RDNA4 hardware (Radeon RX 9070 and RX 9060 series), signaling a shift from broad compatibility toward hardware-specific fidelity. A flagship component, FSR Radiance Caching, leverages an AMD-trained neural model to predict indirect lighting bounces at the second ray intersection to cut GPU work and improve ray-traced global illumination, a move that could narrow DLSS’s quality lead but concentrates benefits on new AMD silicon and changes the competitive and implementation calculus for developers and GPU customers.

Analysis

AMD has announced FSR Redstone, a suite of four ray-tracing and frame-generation technologies (three new, one renamed) intended to raise FidelityFX Super Resolution image quality closer to Nvidia’s DLSS. A highlighted component, FSR Radiance Caching, uses an AMD-trained neural model to predict indirect lighting “as soon as the second ray intersection,” which AMD says reduces GPU work for global illumination. The feature set is explicitly targeted at developers already using FSR 3.1 or FSR 4, but unlike previous FSR releases it will be limited to AMD’s newest RDNA4 GPUs, specifically Radeon RX 9070 and RX 9060 series.

This product strategy represents a shift from broad compatibility toward hardware-specific fidelity; historically FSR ran on a wide range of GPUs while DLSS delivered better image quality but was limited to Nvidia RTX 20-series and newer. Sentiment outputs show a mildly positive market reaction (sentiment_score 0.18, market_impact_score 0.22) with per-ticker signals favoring NVDA (0.5) over AMD (0.2) and INTC (0.0). That suggests the market views Redstone as a constructive technical step but not an immediate displacement of Nvidia’s lead.

Key investor implications are concentrated adoption risk and execution: Redstone can narrow the quality gap if developers integrate it and RDNA4 adoption grows, but its hardware restriction constrains addressable market compared with prior FSR releases. Short-term moves will depend on independent benchmarks, OEM and game-engine uptake, and whether AMD can drive RDNA4 share gains without eroding the advantages of FSR’s prior broad compatibility.

Key Decisions for Investors

• Wait for independent benchmarks comparing FSR Redstone (especially Radiance Caching) to Nvidia DLSS before materially increasing AMD exposure, monitor image-quality and performance data closely
• Track RDNA4 attach rates and initial sales trends for Radeon RX 9070/9060 and the number of game titles adopting Redstone features as the primary adoption signals for upside to AMD, adjust exposure if adoption is weak
• Maintain a cautious overweight or hold on Nvidia given its stronger sentiment signal (NVDA 0.5) and entrenched DLSS lead, but reassess if AMD’s Redstone shows clear, reproducible parity in mainstream titles
• Treat Intel as neutral in this cycle; focus on developer support, driver updates, and any competitive responses from Nvidia or pricing shifts that could change market dynamics