Original Research Index · Sourced Rows

DLSS, FSR & Frame Generation
Laptop Benchmark Database

Search 137 published gaming-laptop results without mixing native rendering, upscaling and generated frames. Every row retains the exact test configuration, game conditions, GPU-power label and original source.

Published Jul 15, 2026Updated Jul 16, 202626 benchmark datasets137 game rows29 gamesNo cross-source composite scores
Generated FPS is not rendered FPS. DLSS Frame Generation, DLSS Multi Frame Generation and FSR Frame Generation can increase displayed frame output by inserting generated frames. Use these rows to study scaling within the same exact test—not as proof of lower latency or a universal laptop ranking.
137Published game rows
26Exact datasets
29Games represented
18Generated-frame rows

Matched Average FPS and 1% Low Charts

The chart selector contains 14 exact-condition groups. Every group keeps the same source dataset, exact laptop, game and published test conditions while the rendering mode changes. Across the full database, 26 rows include a published 1% low and 11 include a minimum FPS.

Loading matched group…
Rendered or upscaled averageGenerated averagePublished 1% low

Generated FPS is displayed output, not a latency or rendered-performance measurement. Performance-per-dollar is intentionally omitted until reliable live prices exist for matching exact SKUs.

Laptop benchmark result database

Loading published benchmark rows…
Exact laptop configurationGame / conditionsRender modeGPU power labelAverage FPS1% lowMinimumPublished source
Loading source database…

Four Evidence Buckets

NativeNative output with upscaling and Frame Generation reported off.
UpscaledDLSS, FSR, XeSS or TSR active while generated frames are off.
GeneratedFG or MFG is explicitly active; displayed FPS includes inserted frames.
UnclearThe publication did not identify enough settings for a safe classification.

How to Compare Rows Correctly

First match the game, output resolution, preset and ray-tracing state. Then match the upscaler, quality mode and Frame Generation multiplier. Finally check the exact laptop configuration, GPU power label and original publication. A high FPS value with MFG x4 does not replace a native or upscaled rendered-performance result.

DLSS Quality vs Balanced: these labels use different internal rendering resolutions, so their FPS and image-quality results are not interchangeable. Compare Quality with Balanced only inside the same exact laptop, game, output resolution, preset, ray-tracing state, DLSS version and Frame Generation state; the database does not apply a universal performance percentage between the two modes.

The “FPS: high to low” control sorts only the currently visible published numbers. It is an exploration tool, not a ranking, because different rows can use different games, settings and protocols. FRAMELIMIT does not average those rows into a synthetic cross-source score.

Can generated FPS be compared directly with native FPS?

No. Generated frames affect displayed output but do not change the underlying rendered-frame rate in the same way. Input latency and image quality require their own measurements.

Why are there only a few clean upscaled-only rows?

Many reviews publish native results or pair upscaling with Frame Generation. Rows enter the upscaled bucket only when the source clearly states that an upscaler is active and generated frames are off.

Why keep rows with unclear settings?

They remain useful as attributed publication records, but excluding them would hide the evidence limitation. The visible Unclear label prevents them from silently entering a native, upscaled or generated comparison.

Continue Research

Razer Blade 14 exact native-game evidence →

Legion 5i related RTX 5070 benchmark evidence →

RTX 50 Laptop TGP Database →RTX vs AMD Technology Guide →Compare Exact Laptop Specs →Benchmark Methodology →