Sony Brings QSSR AI Upscaling to Standard PS5: 2 Games First [2026]

By Indie Kings | October 2, 2026

Updated October 2, 2026: Sony has brought Quick Spectral Super Resolution AI upscaling to the standard PlayStation 5 as a new developer library per VideoCardz by WhyCry on October 1 2026 relaying Sony PlayStation Blog and Digital Foundry. Marvel Wolverine and Ghost of Yotei are the first two games with patches available today per the same relay.

Image: QSSR art. Credit: Sony via PlayStation Blog.

Quick Spectral Super Resolution is now a standard PlayStation 5 feature per VideoCardz by WhyCry on October 1 2026 relaying Sony PlayStation Blog and Digital Foundry. The report describes QSSR as an AI upscaling library for the base console, not a Pro only extra. That framing matters because upscaling libraries shape resolution choices, frame pacing, image stability, and performance headroom for every mode in a game. A new first party upscaler on the base model can affect many future releases if developers adopt it broadly.

Two games lead the rollout per the same VideoCardz relay. Marvel Wolverine and Ghost of Yotei both received patches today that add QSSR support on standard PS5 hardware. Both titles are high profile first party releases with demanding visuals and large audiences. Their selection as launch vehicles signals confidence from Sony in image quality and stability. It also gives Digital Foundry real shipping code to measure instead of a controlled demo.

Broad rollout to PlayStation developers is planned per the VideoCardz relay of Sony PlayStation Blog. The report says games that already use PSSR need little work to move to QSSR. It describes QSSR as close to a drop in replacement with days of retuning work in areas such as texture streaming. That detail suggests Sony designed the library to reuse tuning concepts, integration points, and developer knowledge from the Pro library. Lower porting friction could speed adoption across first party and third party studios [NEEDS VERIFICATION for third party timing].

The thread around this launch is unusual because the same Marvel Wolverine now renders on KytyPS5 in emulation in separate local coverage. That emulation milestone is distinct from official QSSR support and it is not linked here per assignment. Readers should keep the two stories separate in their minds. One story is Sony shipping AI upscaling on real base hardware through official patches. The other story is an unofficial emulator drawing gameplay frames slowly on PC hardware. This article covers only the official QSSR launch except for this clarifying note.

Attribution matters for this article because claims about cost, hardware math, and developer effort travel fast. Every verified core fact in this draft comes from VideoCardz by WhyCry on October 1 2026 relaying Sony PlayStation Blog and Digital Foundry. Figures for frame time, resolution, frame rate deltas, output targets, hardware precision, and porting effort are credited to that chain. Context, explanation, buying implications, and forward looking analysis beyond that relay are marked [NEEDS VERIFICATION] where they go past the cited measurements.

What QSSR Is

QSSR stands for Quick Spectral Super Resolution per the VideoCardz relay of Sony PlayStation Blog. It is described as an AI upscaling library for the standard PS5. An upscaling library reconstructs a higher resolution output from a lower resolution rendered input using temporal data, spatial analysis, and learned image priors. The goal is an image that looks closer to native high resolution while the GPU shades far fewer pixels. Good upscaling buys back frame time for effects, density, stability, or frame rate.

QSSR arrives as part of the PlayStation developer toolkit rather than as a user toggle [NEEDS VERIFICATION for SDK version and menu wording]. Players do not download QSSR separately. Developers integrate the library into a game build, tune it for each visual mode, and ship it inside a patch or a new release. Marvel Wolverine and Ghost of Yotei demonstrate that path because both added QSSR through patches issued today per the VideoCardz relay. Future games can ship with QSSR from day one if studios choose to integrate it during production [NEEDS VERIFICATION for launch timing of the next wave].

The library concept matters more than any single mode name. A library handles inputs such as color, depth, motion vectors, exposure, and history buffers, then produces a reconstructed frame for display [NEEDS VERIFICATION for exact QSSR input set]. Each engine feeds those inputs in its own way, which is why tuning matters per title and per mode. Texture streaming is named in the relay as one area that needs days of retuning when moving from PSSR to QSSR. That clue suggests resolution dependent systems must be rebalanced when the reconstruction target or input range changes. Mip bias, residency windows, and streaming budgets often shift when effective resolution changes.

QSSR is positioned for the standard PS5, which has sold in very large numbers through several hardware revisions [NEEDS VERIFICATION for install base figure]. An upscaler that improves the base console reaches the largest PlayStation audience at once. That reach explains why Sony would invest in a streamlined neural net and a base specific implementation instead of leaving AI upscaling as a Pro showcase. Better reconstruction on base hardware can lift fidelity modes, stabilize performance modes, and extend the useful life of the platform without changing silicon. Developers gain another lever beside dynamic resolution, variable rate shading, and effects scaling [NEEDS VERIFICATION for how QSSR interacts with those systems].

The first two games were chosen with clear intent per the shape of the news. Marvel Wolverine is a new action release built around dense detail, fast traversal, and character close ups. Ghost of Yotei is an open world title with foliage, particles, weather, and long view distances. Both stress temporal stability in different ways. Fast motion tests ghosting and edge persistence. Vegetation and particles test fine detail preservation. Long views test distant aliasing and shimmer. Shipping QSSR in both genres at once gives developers two reference points instead of one narrow demo.

Patch delivery today means base PS5 owners can test QSSR without buying new hardware per the VideoCardz relay. Players need the latest game patch and a standard PS5 console on current system software [NEEDS VERIFICATION for minimum firmware]. No Pro console is required. No paid upgrade is described in the relay. Mode selection still matters because developers choose output targets per mode. A 1440p target behaves differently from a 2160p target in cost and sharpness, as Digital Foundry measurements show. Players should read patch notes and mode descriptions before judging the result from one mode alone.

QSSR does not replace good engine fundamentals [NEEDS VERIFICATION for Sony guidance]. Clean motion vectors, stable temporal history, correct transparency handling, and careful user interface composition still decide final quality. Upscalers amplify both good inputs and bad inputs. A stable base input with accurate motion data reconstructs cleanly. A noisy input with broken motion or flickering effects can smear or crawl after reconstruction. That is why drop in replacement still needs days of retuning. Engineers must verify edges, foliage, hair, particles, water, volumetrics, and interface layers frame by frame.

The name similarity to PSSR is intentional but the targets differ. PSSR is PlayStation Spectral Super Resolution for PS5 Pro. QSSR is Quick Spectral Super Resolution for standard PS5 per the relay. Both use learned reconstruction ideas refined through Project Amethyst collaboration between Sony and AMD. Both aim to turn lower internal resolution into convincing higher output resolution. The difference lies in hardware capability, numeric precision, network size, and per platform tuning. The comparison table below summarizes that split in plain terms.

The table above keeps the marketing separate from the engineering. Quick does not mean low quality by default. It signals a network and implementation tuned for speed on base silicon. Spectral signals continuity with the Pro research line rather than a wholly separate algorithm family. Standard PS5 owners should expect a library tuned around their GPU limits, not a direct copy of Pro settings forced onto weaker machine learning throughput. That platform specific tuning is the reason Sony built a separate library instead of backporting PSSR unchanged.

Developer rollout is described as broad in the relay. Broad likely means documentation, samples, and support through normal PlayStation developer channels [NEEDS VERIFICATION for SDK access and support terms]. Studios with shipped PSSR integrations have the shortest path because much of the data plumbing already exists. Studios on custom temporal upscalers or FSR 3 face a larger evaluation because they must compare image quality, performance, integration cost, and platform exclusivity. First party results in Wolverine and Yotei will serve as sales pitches inside those evaluations. Clean edges plus stable performance in shipping games persuade faster than slide decks.

Players often ask whether QSSR changes art direction or performance targets. An upscaler alone does not add ray tracing, larger crowds, or richer physics [NEEDS VERIFICATION for Sony performance guidance]. It changes the cost of reaching a given output resolution. Studios can spend saved milliseconds on higher frame rates, steadier dynamic resolution, denser effects, or higher settings elsewhere. They can also bank the savings for stability without visible changes. Each studio will choose differently per mode. Patch notes and Digital Foundry analysis remain the best way to see which trade each game made.

Pro Versus Base Tech Split

The technical split starts with Project Amethyst per the VideoCardz relay of Sony PlayStation Blog. Project Amethyst is the collaboration between Sony and AMD around machine learning assisted graphics. PSSR on PS5 Pro came from that work. QSSR is described as derived from PSSR through the same collaboration line, with a streamlined neural net plus a PS5 specific implementation. Derivation here means shared research, shared training ideas, and shared integration lessons, not identical code running on both consoles. The base version is reworked for different throughput and different numeric strengths.

Pro hardware is described as having about 300 TOPS plus INT8 support per the relay. TOPS expresses trillions of operations per second for machine learning workloads. INT8 means 8 bit integer math, which packs more operations into the same silicon budget when accuracy allows. That combination gives the Pro headroom for larger networks, higher output resolutions, and more complex reconstruction passes. Developers targeting Pro can lean on that throughput for premium modes. The relay frames those figures as the Pro advantage that the base console lacks.

Standard PS5 takes a different path with FP16 math plus packed math and GPU optimizations per the relay. FP16 means 16 bit floating point math, which preserves more dynamic range per value than INT8 but costs more per operation. Packed math refers to techniques that process multiple smaller values together to recover throughput on architectures without dedicated low precision accelerators [NEEDS VERIFICATION for exact instruction mapping on base PS5]. GPU optimizations refer to platform specific tuning of shaders, memory traffic, and scheduling to keep the streamlined net fast enough for real game frame budgets [NEEDS VERIFICATION for exact optimization list]. Together they make AI upscaling practical without Pro machine learning blocks.

Streamlined neural net is the key phrase for image quality expectations. A smaller net runs faster but has less capacity to resolve ambiguous detail. A larger net resolves more nuance but needs more time and more machine learning throughput. Sony appears to have searched for a size that fits base frame budgets while retaining visible gains over classic temporal upsampling [NEEDS VERIFICATION for training method and net size]. Platform specific implementation then squeezes the remaining cost through careful coding for the base GPU. That two part strategy is common when backporting a premium feature to a mass market chip.

Precision choice affects more than speed. INT8 quantization can introduce banding or detail loss if calibration is poor, but it unlocks large throughput on supported hardware. FP16 keeps smoother gradients and simpler porting from training, but it needs careful packing and scheduling to hit frame time on base hardware. Sony choice of FP16 with packed math on base PS5 suggests a preference for accuracy and compatibility over peak theoretical throughput [NEEDS VERIFICATION for Sony reasoning]. Developers benefit because FP16 paths are familiar from existing graphics and compute work. Tooling, debugging, and profiling stay closer to known GPU workflows.

Integration friction is low by design per the relay. Games that already use PSSR need little work to adopt QSSR. The relay calls it close to a drop in replacement with days of retuning work such as texture streaming. Drop in means function signatures, resource wiring, and engine hooks stay similar enough that engineers swap libraries without rewriting render passes. Days of retuning means image validation, resolution rebalancing, and streaming adjustments still take real time. Texture streaming is singled out because effective resolution changes alter which mip levels should be resident. Too little streaming work causes soft textures. Too much wastes memory and bandwidth.

That retuning window matters for scheduling. Days per title is short compared with months for a new renderer feature [NEEDS VERIFICATION for typical upscaler integration time]. A short window lets Sony seed QSSR across many Pro supported games quickly. It also lets small teams evaluate without committing a full graphics engineer for a quarter. Larger custom integrations still need quality review across biomes, lighting conditions, weather, crowds, and user interface screens. Fast porting does not mean zero testing. It means testing starts from a working baseline instead of from scratch.

Hardware differences also shape output resolution strategy. Digital Foundry via the VideoCardz relay found 2160p output costs about 4.4 ms while 1440p reconstruction from 864p costs far less in Wolverine testing. That gap reflects compute scaling with output pixels plus network cost per pixel. Base hardware favors 1440p as the better target because the cost fits more easily into 30 FPS and 60 FPS budgets. Pro hardware can reach higher with more comfort [NEEDS VERIFICATION for Pro side by side cost]. Developers choose per mode based on frame budget, display target, and quality bar. There is no single correct output for every game.

Memory and bandwidth shape the result alongside compute [NEEDS VERIFICATION for QSSR memory footprint]. Upscaling passes read history buffers, motion fields, and intermediate features while writing reconstructed output. Higher output resolution increases traffic at every step. Base PS5 unified memory is fast but shared between CPU and GPU, so extra passes compete with streaming, physics, audio, and system work. A streamlined net helps here by touching less data per pixel. Platform specific GPU optimizations likely focus on cache reuse, wavefront occupancy, and async overlap [NEEDS VERIFICATION for exact techniques]. Those details decide whether QSSR fits cleanly or forces cuts elsewhere.

The collaboration angle deserves plain language. Project Amethyst pairs Sony system and game knowledge with AMD graphics and machine learning architecture [NEEDS VERIFICATION for program scope beyond upscaling]. Sony understands console frame budgets, developer workflows, and living room display needs. AMD understands RDNA roadmaps, numeric formats, and training to deployment pipelines. PSSR proved the approach on Pro. QSSR applies the lessons to the larger base install base. Future AMD hardware and future PlayStation hardware may share further advances from the same line, but that is forward looking and marked [NEEDS VERIFICATION for roadmap].

For developers the split creates a clean two target story. Pro gets the larger network with INT8 throughput and higher output headroom. Base gets the streamlined FP16 network tuned for its GPU. Shared concepts and similar integration reduce duplicate work. Per platform tuning handles the rest. That structure mirrors how studios already manage fidelity and performance modes across consoles. One codebase feeds both, with conditional tuning per device. QSSR slots into that model without forcing a new engine architecture [NEEDS VERIFICATION for engine plugin availability].

For players the split means expectations should stay platform specific. Base PS5 QSSR should be judged against prior base upscalers such as Insomniac temporal upscaling and FSR 3, not against Pro PSSR at higher outputs. The relay invites exactly that comparison through Digital Foundry testing in Wolverine and Yotei. Wolverine is measured against the Insomniac upscaler. Yotei is compared with FSR 3 in performance mode. Those are the relevant baselines for base console owners. Cross console comparisons can inform buying choices but they do not measure whether QSSR succeeds on its own target.

DF Numbers

Digital Foundry measurements via the VideoCardz relay give the first hard costs for QSSR in shipping code. In Marvel Wolverine, QSSR adds 1.5 to 1.8 ms over the Insomniac upscaler when reconstructing 864p to 1440p per Digital Foundry via VideoCardz. Milliseconds here mean GPU frame time added by switching upscalers at matched settings. At 60 FPS the full frame budget is about 16.7 ms, so 1.5 to 1.8 ms is a visible but manageable slice. At 30 FPS the budget is about 33.3 ms, so the same cost is a smaller share. Context decides whether the cost is worth the image gain.

The same test reports about 8 to 10 FPS in an unlocked test per Digital Foundry via VideoCardz. Unlocked means the frame rate cap is removed so cost differences show directly as frame rate changes. A drop of 8 to 10 FPS sounds large in isolation, but unlocked deltas exaggerate the feel versus capped modes. Capped 60 FPS mode compensates via dynamic resolution per the relay. Dynamic resolution lowers internal rendering resolution during load to hold the target frame rate. The upscaler then reconstructs the chosen output from that moving input. Net result can be a steadier 60 FPS with slightly softer input in heavy scenes.

Output resolution changes the bill sharply. A 2160p output costs about 4.4 ms per Digital Foundry via VideoCardz. That is roughly two to three times the 1440p path cost in the same Wolverine test. More output pixels mean more network work, more memory traffic, and more display processing. The relay frames 1440p as the better target on base hardware. Developers choose per mode, so a performance mode may favor 1440p for speed while a fidelity mode may still chase 2160p for sharpness on large displays [NEEDS VERIFICATION for mode names in Wolverine patch]. Players on 4K sets will still get a 4K signal through display scaling, but reconstruction quality differs from native 2160p reconstruction.

Input resolution matters as much as output. The measured 864p to 1440p path starts from a modest base and climbs by a factor near 2.8 in pixel count [NEEDS VERIFICATION for exact scale math as reported]. Lower input saves shading cost but gives the network less detail to work with. Higher input preserves more source detail but eats the savings. The 864p choice reflects Insomniac balance for 60 FPS play with dense effects. QSSR must prove it cleans edges, stabilizes speculars, and preserves hair and foliage from that input without smearing. The relay does not provide full image quality scores, only cost and frame rate context via Digital Foundry. Visual judgment still needs direct screenshot and video review [NEEDS VERIFICATION for Digital Foundry quality verdict wording].

The Yotei result is simpler on the surface. Roughly same frame rate as FSR 3 in unlocked performance mode per Digital Foundry via VideoCardz. Parity on speed means studios can consider QSSR without conceding performance in that mode. Image quality then becomes the decider. Does QSSR resolve foliage, resolve distant geometry, handle particles, and hold edges better than FSR 3 at matched input. The relay does not publish a full quality verdict in the excerpt. Players should wait for direct comparisons before declaring a winner on looks [NEEDS VERIFICATION for side by side stills]. Speed parity alone does not prove visual superiority.

Dynamic resolution deserves extra explanation because it changes how QSSR costs feel. A fixed resolution test shows clean deltas but not shipping behavior. Shipping 60 FPS modes often float internal resolution within a window to protect frame pacing. When QSSR adds 1.5 to 1.8 ms, the dynamic system can drop input slightly to recover the budget. The output stays at the chosen target while the input softens for brief moments. Most players prefer that trade to stutter or tearing. The relay explicitly says 60 FPS mode compensates via dynamic resolution, which signals Insomniac tuned the window with QSSR in mind.

Display size and viewing distance change the value of 1440p versus 2160p [NEEDS VERIFICATION for viewer study data]. On a large 4K display viewed close, 2160p reconstruction can look cleaner on fine lines and distant detail. On a smaller display or from couch distance, well reconstructed 1440p can look close while running faster. Base PS5 owners with 1440p displays get a natural fit because output matches panel resolution without extra scaling. Owners with 4K displays still benefit because the television scales a clean 1440p reconstruction, though not identically to a 2160p reconstruction. Personal setup decides how much the 4.4 ms path is worth.

Testing limits should stay in view. The relay summarizes Digital Foundry findings but not full methodology in the excerpt. Readers do not see exact scenes, run counts, error bars, firmware, or capture hardware from the summary alone [NEEDS VERIFICATION for full Digital Foundry methodology]. Unlocked deltas vary by scene density, effects load, and camera motion. Dynamic resolution behavior varies by mode tuning. One game per data point is a start, not a library wide proof. Wolverine numbers describe Insomniac engine and content. Yotei parity describes Sucker Punch engine and content in one mode. Other engines will report different costs.

Cost per pixel is not the whole image quality story. A slower upscaler can still win if it removes shimmer, crawling, ghosting, or breakup that cheaper methods leave behind. A faster upscaler can still lose if it smears fine detail or lags on motion. Digital Foundry cost reporting answers whether QSSR fits frame budgets. Visual verdicts need pixel level review across motion, not just still frames. Readers should look for hair against bright skies, foliage in wind, rain and particles, specular highlights on metal and water, thin wires and fences, and text on signs during camera moves [NEEDS VERIFICATION for which scenes Digital Foundry judged]. Those cases expose temporal weaknesses quickly.

For base PS5 owners the practical read is grounded. QSSR costs more than the custom Insomniac upscaler in Wolverine at 1440p, but dynamic resolution absorbs it in capped 60 FPS play per the relay. QSSR at 2160p costs substantially more and fits fewer budgets, which is why 1440p is framed as the better target. QSSR matches FSR 3 on speed in Yotei unlocked performance mode per the relay. Those three sentences summarize the verified performance picture. Everything beyond them about cleaner edges or broader wins awaits direct visual proof and wider game coverage [NEEDS VERIFICATION for quality ranking].

What Comes Next

Broad rollout to PlayStation developers is planned per the VideoCardz relay of Sony PlayStation Blog. That phrasing points to documentation, SDK access, and support rather than a single surprise drop. Studios already shipping PSSR on Pro have the fastest lane because QSSR is described as near drop in with days of retuning. Texture streaming is named as one retuning area. Resolution dependent effects, history handling, and user interface compositing likely need review as well [NEEDS VERIFICATION for official porting guide]. Expect first party studios to move first because they have direct access and shared technology.

PSSR games are the natural porting pool per relay logic. Any title that already wires motion vectors, history buffers, and reconstruction hooks for PSSR can reuse much of that work for QSSR. Days of retuning is short enough to fit into a regular patch cycle. That speed could produce a steady drip of QSSR patches across late 2026 and 2027 [NEEDS VERIFICATION for patch calendar]. Players should watch patch notes for QSSR mentions alongside dynamic resolution tweaks and mode rebalancing. A patch that adds QSSR may also adjust internal resolution windows to keep frame targets stable.

New releases can integrate QSSR from the start of production [NEEDS VERIFICATION for engine plugin status]. Early integration lets artists and engineers tune content with reconstruction in mind instead of retrofitting. Foliage density, hair card design, particle sizes, and specular intensity all interact with upscaling. Building with QSSR targets early can reduce late surprises. Middleware engines may add plugins or samples to lower the bar further [NEEDS VERIFICATION for Unreal and proprietary engine plans]. Until then, custom integration by experienced graphics teams will lead.

Third party adoption will hinge on cost versus reach [NEEDS VERIFICATION for licensing and support terms]. QSSR helps only PlayStation builds, while FSR, XeSS, or custom temporal upscalers can span PC and other consoles. Multiplatform teams must weigh a PlayStation specific win against maintenance of another code path. If QSSR clearly beats cross platform options on image stability at similar cost, many will take it for the PS5 build. If the visual gap is small, some will stay with shared solutions to save test time. Yotei parity with FSR 3 on speed is a useful data point, but visual comparisons will close the argument.

Mode design may evolve as developers learn the cost curve. The relay establishes 1440p as the sweet spot and 2160p as costly on base hardware. Studios could standardize on 1440p reconstruction for 60 FPS performance modes and reserve 2160p reconstruction for 30 FPS fidelity modes [NEEDS VERIFICATION for upcoming mode plans]. They could also offer user choice between sharper output and steadier input resolution. Dynamic resolution windows will likely widen slightly where QSSR adds cost, then tighten as tuning improves. None of that changes gameplay systems. It changes how clean the image looks during motion.

Display trends support the 1440p focus [NEEDS VERIFICATION for market data]. More monitors and televisions now accept 1440p cleanly, and many 4K sets scale 1440p well for games. A strong 1440p reconstruction can look excellent in motion while leaving headroom for effects and stability. The 2160p path remains for showcase modes, photo modes, and slower paced titles where frame budget allows [NEEDS VERIFICATION for title plans]. Developers choose per mode, so players will see both strategies. Reading mode tables before choosing fidelity or performance remains good practice.

Project Amethyst work continues in the background [NEEDS VERIFICATION for roadmap]. QSSR proves the research scales down to base hardware, not only up to premium hardware. Future advances could improve quality at the same cost, reduce cost at the same quality, or extend to higher outputs on future chips. AMD benefits because console lessons feed PC and integrated graphics upscaling. Sony benefits because a shared research line serves two console tiers at once. Players benefit when better reconstruction arrives as a software patch rather than a hardware purchase. That software leverage is the strategic point of QSSR.

Competition will not stand still [NEEDS VERIFICATION for rival roadmaps]. FSR continues to evolve on PC and consoles. Custom temporal upscalers at first party studios keep improving with game specific tuning. Machine learning approaches on other platforms push quality higher each year. QSSR must earn its place title by title through measurable stability and reasonable cost. The Wolverine and Yotei launches are opening bids, not final judgments. Broader testing across engines, biomes, and motion patterns will decide reputation.

What should players watch next. Watch patch notes for Marvel Wolverine and Ghost of Yotei to confirm mode targets and dynamic resolution behavior on your display [NEEDS VERIFICATION for patch note text]. Watch Digital Foundry full analysis via VideoCardz relay for visual verdicts beyond cost. Watch PlayStation developer channels for SDK notes and sample updates [NEEDS VERIFICATION for publication timing]. Watch third party announcements for the first non Sony studio to commit. Each signal marks adoption moving from two showcase patches to a platform wide option. Until then, base PS5 owners can enjoy two early examples while waiting for wider support.

A final note keeps coverage clean. The same Marvel Wolverine now renders on KytyPS5 in emulation per separate local coverage that is not linked here. That result belongs to unofficial PC emulation and it runs slowly. It does not reflect official QSSR quality, official performance, or official support. Official QSSR runs on real PS5 hardware through licensed game patches. Unofficial emulation runs through translation layers on PC hardware with many faults. Do not compare frame rates or image quality across those worlds. Judge QSSR on base PS5 hardware with patched games and measured output.

FAQ

What is QSSR on standard PS5?
Per VideoCardz by WhyCry on October 1 2026 relaying Sony PlayStation Blog, QSSR or Quick Spectral Super Resolution is an AI upscaling library for the standard PS5 that reconstructs higher output resolution from lower internal rendering.

Which games get QSSR first?
Per the same VideoCardz relay, Marvel Wolverine and Ghost of Yotei are first with patches available today on standard PS5 hardware.

How much does QSSR cost in Marvel Wolverine?
Per Digital Foundry via VideoCardz, QSSR adds 1.5 to 1.8 ms over the Insomniac upscaler when reconstructing 864p to 1440p, with about 8 to 10 FPS difference in an unlocked test and compensation via dynamic resolution in capped 60 FPS mode.

Is 2160p or 1440p better for QSSR on base PS5?
Per Digital Foundry via VideoCardz, 2160p output costs about 4.4 ms so 1440p is the better target on base hardware, and developers choose the output per mode.

How does QSSR compare to FSR 3 in Ghost of Yotei?
Per Digital Foundry via VideoCardz, QSSR runs at roughly the same frame rate as FSR 3 in unlocked performance mode, with visual quality comparison still awaiting full review [NEEDS VERIFICATION for quality verdict].

Do PSSR games need major rework to use QSSR?
Per VideoCardz relaying Sony PlayStation Blog, PSSR games need little work because QSSR is close to a drop in replacement with days of retuning such as texture streaming, and broad rollout to PlayStation developers is planned.

Bottom Line

Sony has moved AI upscaling down the stack with QSSR for standard PS5 per VideoCardz by WhyCry on October 1 2026 relaying Sony PlayStation Blog and Digital Foundry. Marvel Wolverine and Ghost of Yotei prove the library in shipping patches today. Wolverine measurements show 1.5 to 1.8 ms over the Insomniac path at 864p to 1440p, about 8 to 10 FPS in unlocked testing, capped 60 FPS compensation through dynamic resolution, and about 4.4 ms at 2160p output with 1440p favored. Yotei shows rough speed parity with FSR 3 in performance mode. Porting looks fast for PSSR titles with days of retuning around texture streaming. The tech split is clear with Pro leaning on about 300 TOPS plus INT8 while base uses FP16 with packed math and GPU optimizations from a streamlined net born of Project Amethyst work with AMD. Broad developer rollout is planned and developers choose outputs per mode. Base PS5 owners get a real new lever without new hardware. Watch full visual analysis and patch notes for mode by mode results [NEEDS VERIFICATION for full quality rankings].

Sources

VideoCardz reporting by WhyCry on October 1 2026 relaying Sony PlayStation Blog and Digital Foundry, as cited throughout. Sony PlayStation Blog via VideoCardz for QSSR library description and developer rollout. Digital Foundry via VideoCardz for Wolverine and Yotei measurements. More coverage at VideoCardz, PlayStation Blog, and Digital Foundry. No other sources used.