Pick your starting point

First check PC Requirements and your current graphics driver. If the game fails to launch or an error blocks play, follow Launch Troubleshooting. Once you can play, use the comparison loop below.[3][4]

A settings experiment you can repeat[1][3]
  1. 01 · RECORD

    Save your baseline

    Note resolution, preset, upscaler and frame generation. Capture the benchmark result and choose a repeatable checkpoint.

  2. 02 · CHANGE

    Move one control

    Change a single option. Re-run the benchmark and the same movement or combat route.

  3. 03 · KEEP

    Judge the whole scene

    Compare dips, image clarity and how aiming feels. Restore the baseline if the change makes your session worse.

Low Core Mode: compare before enabling

In Tom’s Hardware’s 1080p High tests with an RTX 5090, Low Core Mode lowered average FPS on all 25 desktop CPUs without improving 1% lows. These three results are examples, not a forecast for your PC.[1]

Three desktop CPU comparisons; the accessible table below gives all six measured FPS values.
Wiki plot of three published FPS pairs. Data: Tom’s Hardware / Jake Roach, October 1. 1080p · High · RTX 5090.[1]
The same data in text — average FPS[1]
Desktop CPUMode offMode onChange
Core i3-1410090.179.2−12.1%
Ryzen 5 7600X118.790.2−24.0%
Ryzen 7 9800X3D168.7133.1−21.1%

An artificially limited two-core test improved with this mode. Mobile CPUs and handhelds were not tested, so these desktop results do not settle every device’s best choice.[1]

For a desktop, start with the mode off and compare if you have a reason to try it. On a handheld, keep the power profile and scene consistent while comparing both states. Keep the choice that improves your actual session.[1]

An RTX 3060 starting profile

Published test result — a starting point to verify[2]
OptionTested choice
Output resolution1920 × 1080
Graphics presetMedium
UpscalingDLSS Quality
Reported resultAbove 60 FPS average in the reviewer’s test

Jeffrey Kampman’s test found that raising this profile to High worsened 1% lows. On the tested RTX 3060, Ludicrous roughly halved performance versus Ultra. Medium also retained more shadow detail than Low in his visual comparison.[2]

Treat the profile as your first trial. Your CPU, scene and game version can change the result. If you use different hardware, start with a comfortable preset and apply the same measurement loop rather than copying an FPS promise.[2]

Follow the symptom you can see

Editorial experiments — change one option at a time[1][2][3]
What happensTry nextWatch for
The whole scene runs slowlyTry the next lower preset, then separately compare a supported upscaler’s Quality mode.Average FPS, dips and clarity while moving.
A lower preset barely changes FPSCompare Low Core Mode states and inspect CPU/GPU readings in the same scene.Whether the measured bottleneck changes; one reading alone is inconclusive.
Still frames look good; movement looks unstableCompare upscaling choices while panning past foliage, wires and bright lights.Flicker, trails, noise and target visibility.
The FPS number rises, but aiming still feels sluggishIf available on your hardware, compare frame generation on/off with the same base settings.Input feel alongside the measured base frame rate.

Can I turn ray tracing off to recover performance?

The game requires hardware ray tracing. The developer’s published specifications are the reference for supported hardware; the launch graphics test also reports no RT-off path.[2][3]

Why can frame generation disagree with the performance overlay?

The launch overlay excludes generated frames. Keep base FPS and generated output separate in your notes, and compare aiming feel as well as the displayed number.[5]

Your next test session

This is a personal test checklist, not a game unlock. Progress is saved in this browser. Keep your settings and FPS values in your own notes or screenshots so you can compare later.

0 / 5 completed

Your checks are saved in this browser. This checklist does not sync with the game.

One complete comparison loop
DoneStepRecord or compare
BaselineCPU/GPU, driver, game version, preset, resolution, upscaler and frame generation.
BenchmarkRun the built-in benchmark; record the result before changing settings.
Real checkpointReplay one fixed route with combat or movement, then capture dips and image issues.
Single changeRepeat with one changed setting and restore the baseline to check the difference.
Keep a profileSave a screenshot of the settings you prefer and test a second representative scene.

The CPU reviewer’s benchmark ran slightly slower than two played scenes. Check your actual route as well.[1]

Sources & further reading

  1. Original tests of 25 desktop CPUs and Low Core ModeTom’s Hardware · Jake Roach · October 1, 2026
  2. Original graphics tests of 43 GPUs and RTX 3060 preset scalingTom’s Hardware · Jeffrey Kampman · October 1, 2026
  3. E-Day final PC specifications and September 29 pre-installXbox Wire / The Coalition · September 17, 2026; specifications image updated September 28
  4. GeForce Game Ready 617.14: E-Day, DLSS and RTX Mega GeometryNVIDIA · September 22, 2026
  5. Emergence Begins: Season 1, launch changes and known issuesThe Coalition · September 30, 2026