LockerGnomeTech news and nostalgia for today's technology professionals
Subscribe

Search

GamingWindowsVALORANTOpen Source

Pitstop checks game stutters before changing Windows, then logs the fixes for undo

Intel’s tracing tools and Riot’s firmware requirements explain why smoother play can require diagnosis before another purchase. The crucial comparison is the same workload twice.

Chris Pirillo kneels beside a race-liveried PC tower in a pit box, timing it with a stopwatch before touching anything
Illustration: LockerGnome

A game booster can’t move the monitor cable out of the motherboard port. Pitstop at least knows when to tell you to do that. Its maker’s examples include the wrong display connection, a monitor stuck at 60 Hz and a PC with one memory stick. Some fixes require getting behind the desk, that glamorous part of PC gaming the screenshots leave out.

The project’s README describes a 60-second measurement using Intel’s PresentMon and labels its 1% low result “FPS in fights.” It also checks latency, jitter and packet loss. Software changes go into an undo log, with a Windows restore point created first. Its maker says Defender, Windows Update and Memory integrity stay enabled.

Those commitments make diagnosis worth separating from the usual promise that a large Fix button can improve every machine. The project explicitly allows for finding nothing wrong. A working PC doesn’t need an invented problem just to justify a download, and somebody deciding whether to spend money on more memory needs to know which limitation the measurement actually found.

Intel’s PresentMon repository explains the measurement machinery underneath. Its tools trace CPU, GPU and display frame durations and latencies across graphics interfaces including DirectX, OpenGL and Vulkan. The collection library analyzes Windows event-tracing events, and its console application can export CSV data for a target application. Pitstop packages that sort of evidence into a result a player can read.

Frame duration is useful because a single average can conceal uneven delivery. As an illustrative calculation, sixty evenly spaced frames in one second arrive about 16.7 milliseconds apart. Delivering the same count with a long pause among them preserves that one-second average while changing what you see. That example explains why the slow part of a run deserves attention alongside the larger number in an FPS counter; it is arithmetic, not a Pitstop benchmark.

A game that won’t start needs a different investigation. Riot’s VAN 9001 support page says that code means Vanguard requires TPM 2.0 but doesn’t detect it enabled. Riot directs the player to its TPM configuration guide before trying again. More graphics performance won’t resolve a prerequisite that prevents the game from launching in the first place.

Pitstop’s own documentation also identifies the limits of its demonstration. The banner shows an example improving from 118 to 231 FPS, rather than a promised result for your machine. Network server-city tests use public stand-in hosts, so their ping may differ from the game. The portable build targets Windows 10 and 11 on x64 and is not yet code-signed. These are the maker’s published capabilities and limitations, not an independent hands-on test here.

A useful next step for tools like this would be preserving enough detail to reproduce a before-and-after run, including which problem was found and which change addressed it. A shareable gain is more persuasive when another person can repeat the workload. The unresolved question for your PC is whether its particular interruption survives the same minute of play after the specific fix.

More Gaming

Every Gaming story →