Active window priority

Settings›Tweaks›Active window priority

Recommendation

Situational

On client Windows 10 and 11, keep Windows default. Check 0x1A only under reproducible full CPU load.

The historical BoosterX test did not reveal a universally better state: default showed a slightly lower average click-to-photon latency, while 0x1A matched the best FPS figures in one CPU scenario without independent repeats.

Usually chosen

Windows default

Main effect

CPU quantum policy

FPS and latency

Depend on load

In short

Win32PrioritySeparation sets the foreground boost and part of the rules for choosing Windows scheduler quanta. They determine how processor time is distributed among ready threads. The setting by itself does not raise a game’s priority or the CPU’s power; an FPS gain is not guaranteed.

For most client computers running Windows 10 or 11, we recommend Windows default (0x02). The 0x1A state switches the fields under test to long fixed quanta with maximum foreground boost. It may be useful only in the narrow scenario where a game reproducibly hits a CPU limit and a comparison confirms an improvement without worsening latency or rare frames.

Windows Server uses different scheduling rules and was not tested in this study. On a server system, keep the original state: the 0x02 recommendation from this article does not apply to it.

What changes

Microsoft documents the lower two bits of the parameter as ForegroundApplicationBoost: 0 means no boost, 1 minimum, 2 maximum. Historical Windows Internals documentation describes the remaining fields in use as the choice of short or long and variable or fixed quanta.

0x02 leaves the choice of quanta to the standard client Windows policy and keeps maximum foreground boost. 0x1A sets long fixed quanta with the same maximum boost. A longer quantum can reduce the frequency of thread switches under heavy CPU load, but at the same time change the responsiveness of competing tasks.

Registry changes

The setting is available in the free version of BoosterX. The table lists its registry values.

Field	Value
Hive and path	HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\PriorityControl
Value name	Win32PrioritySeparation
Type	REG_DWORD
Windows default	0x02 (2), the documented initial value for client Windows
Situational state	0x1A (26)
Revert	0x02 (2)

Use BoosterX to change and revert it. There are no ready-made .reg or PowerShell commands on this page.

On the tested Windows 11 25H2, the value was set explicitly and equaled 2. Therefore, the absence of the DWORD cannot be considered the factory state for all Windows installations. If the value is absent, the parameter is not explicitly set in the registry. This differs from a recorded value of 0. There is no need to create an entry with a random number.

Value calculator

6-bit decoder

Real mode check

Win32PSCalculator ↗

Enter the value you found in the tweak list. The calculator will show only the six bits used and the canonical combination with the same mode. It does not change anything on your computer.

Registry value

Parse

Hex with 0x prefix, otherwise decimal.

Default 0x02Explicit default 0x26Contextual 0x1A

Equivalent mode0x26

Windows default: on client Windows, corresponds to explicit mode 0x26.

000010

Entered

0x00000002 · 2

After 0x3F mask

0x02 · 2

Quantums

System default → short

Type

System default → variable

Foreground boost

Maximum · 3:1

The calculator follows the interpretation of the open-source project Win32PSCalculator: it takes into account only the lower six bits and shows the canonical value of the same mode. For example, Windows default 0x02 and the explicit 0x26 correspond to short variable quanta with maximum foreground boost on client Windows. This explains why setting 0x26 on top of default usually does not create a new mode.

The calculator shows which mode a bit combination sets. It does not predict FPS or latency. Different numbers can set the same mode.

When to check 0x1A

• the game or workload consistently keeps the CPU near full load;
• frametime worsens under full CPU load;
• you can run several identical paired runs;
• not only average FPS is compared, but also P1, P0.1, the frametime graph and input latency.

Do not choose 0x1A just because someone else’s settings profile uses it. If the CPU is not fully loaded, there may be no difference. A Windows 10 result also cannot be automatically carried over to Windows 11.

What BoosterX measurements showed

In the historical Valorant series on Windows 10 22H2, 300 click-to-photon measurements per value were used at a recorded 100% CPU load.

Metric	Windows default 0x02 / explicit equivalent 0x26	0x1A
Click-to-photon AVG	15.28 ms	15.68 ms
Click-to-photon SD	3.12 ms	3.31 ms
FPS AVG	334.4	355.1
P1	111.6	256.0
P0.1	36.9	41.0

The difference in average latency was 0.40 ms in favor of default, with a spread of about 3 ms. In the saved record, FPS was higher for 0x1A. However, only one FPS run was saved for each state, so the repeatability of this result has not been confirmed.

The full matrix of values, methodology and limitations are given in the study “Win32PrioritySeparation: latency and FPS under full CPU load”.

How to verify the result

1. Save Windows default and fix the same game scene, FPS limit, power and background load.
2. Perform several runs with the original setting and save frametime, P1 and P0.1.
3. In BoosterX, open “Optimization” → “Tweaks” → “Scheduler and memory” → “Win32PrioritySeparation”, select 0x1A, click “Apply” and perform the suggested restart.
4. Repeat the same runs in shuffled order.
5. Keep 0x1A only if the target metric improves repeatably without worsening latency, rare frames and background tasks.

How to revert

In BoosterX, open “Optimization” → “Tweaks” → “Scheduler and memory” → “Win32PrioritySeparation”, select “Default” and click “Apply”. If “Instant apply” is enabled, a separate click is not required. Perform the suggested restart and make sure the value is again equal to 0x02.

Sources and scope of verification

• Win32_OperatingSystem.ForegroundApplicationBoost, Microsoft Learn - the relationship with the Registry and foreground boost values.
• CPU Analysis, Microsoft Learn - the role of quantum, priority and switches in scheduler operation.
• Scheduling Priorities, Microsoft Learn - priorities, preemption and round-robin thread scheduling.
• Context Switches, Microsoft Learn - how context switching works.
• Windows Internals Sixth Edition sample chapters, Microsoft Press - the historical description of the bit fields and client quantum policy.
• Public BoosterX measurement table - historical click-to-photon and FPS aggregates.
• Win32PSCalculator - an open-source desktop calculator of equivalent parameter values.

The behavior of all quantum variants was not tested on every Windows build. The effect of 0x1A on Windows 11 and modern gaming systems has not yet been reproduced. The recommendation does not apply to Windows Server.

Last verified: 2026-09-20.