PC optimizer to boost FPS: which changes can move your frame rate

What an optimizer can and cannot do for FPS

Frame rate is decided by how much work the game asks of your processor and graphics card and how quickly they get through it. Windows sits between the game and that hardware, so its settings can only do two things for frame rate: stop other work competing for the same resources, and stop power or scheduling policies holding the hardware back. Neither makes the hardware itself faster.

That is why the same optimizer can appear to transform one PC and do nothing on another. A machine with a dozen launchers, an overlay recording in the background and a laptop power profile has a lot in the way. A clean desktop already running flat out on its graphics card has almost nothing left to remove. When a game is limited by the graphics card, Windows tweaks rarely change the average at all, although they can still make frame delivery more even.

So the honest way to think about an optimizer is as a way of clearing obstacles, and the useful question is which obstacles a given category of change actually touches. That is what the rest of this page works through. It makes no promise of a figure, because any figure measured on one PC would not transfer to yours.

Categories of change, sorted by their link to frame rate

Every optimizer's settings fall into a handful of categories. They are not equally connected to frame rate, and several popular ones are not connected to it at all.

Read the table from the top down when deciding what to test. The first two rows are where a genuinely cluttered or power-limited PC has room to change. The last three are where most of the confident claims in this category turn out to be empty.

Power plans and processor states

Windows' Balanced plan is designed to save energy, which means letting processor cores park when idle and stepping them down to lower performance states between bursts of work. A game produces uneven load, and on some systems the delay in waking a parked core or raising its state shows up as uneven frame pacing, particularly in games that lean heavily on the CPU.

Choosing the High Performance plan, or raising the minimum processor state, keeps cores ready. On a desktop this costs some idle power. On a laptop it costs battery life and heat, and heat can undo the gain by forcing the chip to throttle. Windows 11 also layers a Power mode slider on top of the Balanced plan, which is worth checking before changing plans at all.

The registry values involved, their defaults and what each one does are covered in core parking and processor performance states, and laptop-specific behaviour is in power throttling.

Background process suspension

Several free tools, Razer Cortex among them, pause or close background processes when a game starts and bring them back when it ends. The effect depends entirely on what was running. On a PC with a browser playing video, a cloud sync mid-upload and three launchers checking for updates, freeing that processor time can matter. On a PC that was already quiet, there is nothing to free.

It matters most on processors with fewer cores and in games that are limited by the CPU rather than the graphics card. You can do the manual version for free: remove unneeded entries under Startup apps in Settings, close overlays you do not use, and check Task Manager during a match to see what is actually taking processor time. Windows 11's Efficiency mode in Task Manager can also push a specific background process to low priority without closing it.

Background capture deserves its own mention. If the Xbox Game Bar is recording the last few minutes of play in case you want to save them, the graphics card is encoding video the whole time. Game DVR and fullscreen optimizations explains where that setting lives and what switching it off changes.

GPU scheduling: worth testing, never worth assuming

Hardware-accelerated GPU scheduling moves management of the graphics card's work queue from the processor onto the card. It arrived in Windows 10 version 2004 and needs a supported card and driver. Microsoft's own announcement said most users should not expect to notice a significant change, because games already hide most scheduling overhead through buffering.

In practice it helps on some combinations of hardware, driver and game and makes things slightly worse on others, which is exactly the kind of setting where the only useful answer comes from testing both states in the games you actually play. Some features, such as frame generation in recent graphics drivers, require it to be on, which can settle the question for you. The value, its registry location and how to switch it are on the hardware-accelerated GPU scheduling reference.

Network and ping tweaks change latency, not frame rate

This is the distinction most optimizer pages leave out, and it is the one that saves the most wasted effort. Three different things get called "lag", and they have different causes:

• Frame rate is how many images your PC produces each second. It happens entirely inside your machine.
• Input latency is the delay between your hand moving and the result appearing on screen. Frame rate, the display and the input device all contribute.
• Ping is the round trip between your PC and the game server. It depends on distance, routing and your connection.

Network settings such as Nagle's algorithm, TCP acknowledgement frequency or the network throttling index change how your PC sends and receives packets. At best they affect latency and stability of the connection. They cannot add a single frame, because frames are not produced by the network stack. Most fast-paced games also send their gameplay traffic over UDP, which the TCP-specific settings do not touch.

If your complaint is that the game feels delayed while the frame counter looks healthy, look at the network and input side. Nagle's algorithm and NetworkThrottlingIndex explain what those two settings genuinely do, and how to fix high ping on a PC starts from the symptom instead.

Frame rate caps and sync settings

One of the changes most reliably felt as smoothness is not an optimizer setting at all. When a game produces frames faster than the monitor can show them, or swings between high and low rates from one moment to the next, delivery becomes uneven. Capping the frame rate a little below your monitor's refresh rate, with the game's own limiter where it has one, keeps delivery steady and leaves the graphics card spare capacity for demanding moments.

On a variable refresh rate monitor, a cap just below the refresh rate also keeps the game inside the range where G-SYNC or FreeSync stays active. Low-latency features in the graphics driver, and NVIDIA Reflex in the games that support it, are designed to work alongside that. None of this raises the average, and a cap deliberately lowers the ceiling, but it is often what turns a game that measures well into one that also feels good. The variable refresh rate reference explains how those pieces fit together.

Settings that look like FPS tweaks but are not

Some of the most copied gaming tweaks sound like they should raise frame rate and are documented by Microsoft as doing nothing of the sort. The multimedia scheduler values in particular circulate in almost every tweak pack, and several of them are either clamped back to the default by Windows or described in Microsoft's reference as unused.

Memory cleaners are a related case. Windows already uses spare memory for caching and hands it back instantly when a program needs it, so forcibly emptying that cache mostly means Windows has to fill it again. None of this shows up as more frames.

We publish the evidence for each of these, value by value, in registry tweaks that do nothing, and the sceptic's version of this whole question is do PC optimizers really boost FPS at all.

How to test whether an optimizer actually helped

A result you did not measure is a feeling, and feelings about frame rate are unreliable in both directions. This method takes an evening, needs only free software, and gives you an answer you can trust for your own machine.

1. Pick a scene you can repeat exactly. A built-in benchmark, a replay, a bot match on a fixed map or a set route through a level all work. A live match does not, because no two are alike.
2. Hold everything else still. Keep the same resolution, graphics preset, driver version and open applications for every run, and restart the PC before each set so leftover background work does not skew one side.
3. Capture frame times, not a counter. Use PresentMon, CapFrameX or the game's own benchmark output and record at least three runs. An on-screen FPS number hides the stutter that frame-time data shows.
4. Change a single thing. Apply one optimizer category, or one individual setting, and nothing else between the two sets of runs.
5. Run the same scene the same number of times. Compare the averages and the slowest frames across both sets. If the difference is smaller than the spread between runs of the same setup, it is noise rather than an improvement.
6. Decide, then move on. Keep the change only if it held up across every run, reverse it if it did not, and only then test the next category.

The same principles govern anything this site says about performance; they are written out in full on how we measure performance. When you know which category helped, the tweaking app comparison shows which tools let you apply just that one.

Where JINSHI PC Tweaks fits

JINSHI PC Tweaks is our own product, so read this section with that in mind. Of its 122 changes, the ones connected to frame rate sit in the power and gaming categories: the High Performance power plan, processor throttling and state settings, background recording, Game Mode and a GPU scheduling toggle. Its network category is a latency and connection category, and we describe it that way rather than as an FPS feature.

It does not promise a frame-rate gain, and nothing on this site quotes one as an expected result. What it offers is those categories applied together, each change listed with its registry path on the full tweak list, and each one reversible on its own, which suits the one-change-at-a-time testing described above. It has no free tier, so the free route on this page is worth trying first. If you decide it is worth it, the JINSHI Tweaks homepage lists what PC Tweaks and the other three apps include.

For a broader look at which kind of optimizer suits a gaming PC, see the best PC optimizer for gaming, and how to choose a PC optimizer for the checklist version.