Recurring ~315-second FPS drops / massive soft page-fault bursts
I am experiencing a highly repeatable performance issue in Apex Legends where the game runs normally at my FPS cap, then approximately every 5 minutes 15 seconds (~315 seconds) FPS suddenly drops for roughly 10–15 seconds, then returns to normal.
This happens regardless of whether I am in combat. It also occurs in the Firing Range, so it is not related to network load, specific map locations, or large fights.
System:
- Intel Core i5-14400F
- NVIDIA RTX 4060
- 16 GB DDR5-4800
- Windows 11
- Apex running through Steam
- DX12 (r5apex_dx12.exe)
Measured behavior
I recorded the issue using both HWiNFO and Windows Performance Monitor.
Example drop timestamps from one session:
- 20:23:17
- 20:28:31
- 20:33:46
- 20:39:01
- 20:44:18
Intervals were approximately:
- 314.7 seconds
- 314.9 seconds
- 314.6 seconds
- 316.7 seconds
So the issue is extremely periodic and repeats roughly every 315 seconds.
During the FPS drop:
- FPS falls from approximately 180 FPS to 130–150 FPS
- CPU/GPU temperatures remain normal
- CPU clocks remain normal
- CPU package power remains normal
- GPU clock remains normal
- GPU is not thermal or power throttling
- RAM usage does not suddenly max out
- VRAM does not max out
- SSD activity does not spike significantly
- GPU workload falls because the GPU appears to wait for the CPU/game
- CPU frame time increases noticeably
PerfMon finding
The most significant finding is that r5apex_dx12.exe generates an enormous periodic burst of page faults at the exact same time as the FPS drops.
During normal gameplay:
~200 page faults/sec
During the problem:
~200,000–290,000 page faults/sec
The vast majority appear to be soft/transition page faults rather than hard disk page faults.
Example second capture:
- Apex page faults during event: ~207,000/sec average, ~225,000/sec peak
- Another event: ~242,000/sec average, ~259,000/sec peak
- Transition Faults/sec: approximately 195,000–229,000/sec
- Working Set changes only slightly
- Private Bytes remain roughly stable
- Virtual Bytes remain roughly stable
- Disk page reads remain very low
This suggests Apex is repeatedly re-accessing/remapping memory already present in RAM rather than suffering normal disk/pagefile thrashing.
WPR / WPA investigation
I also captured a Windows Performance Recorder trace and inspected it with Windows Performance Analyzer.
A large VirtualAlloc/commit path resolves through:
ntdll.dll → kernel32.dll → r5apex_dx12.exe
The major allocation branch stays inside r5apex_dx12.exe.
Easy Anti-Cheat activity in that particular VirtualAlloc stack was negligible compared with the Apex allocation branch.
Unfortunately Apex private symbols are unavailable, so I cannot resolve the exact internal Apex function beyond r5apex_dx12.exe.
Troubleshooting already performed
I have already tested or ruled out:
- Steam file integrity verification
- complete Apex reinstall
- switching from EA App version to Steam version
- deleting/resetting Apex configs
- clearing Apex caches
- low/default graphics settings
- different FPS caps
- fullscreen optimizations
- Windows Game Mode on/off
- HAGS on/off
- NVIDIA Reflex variations
- NVIDIA Threaded Optimization disabled
- multiple NVIDIA/game settings
- CPU affinity / P-core testing
- Process Lasso affinity/hyperthreading tests
- Process Lasso SmartTrim — disabled
- Process Lasso Watchdog rules — none
- scheduled Windows tasks — no ~5-minute repeating task found
- USB selective suspend — already disabled
- Realtek Ethernet device/driver — disabled/removed
- Easy Anti-Cheat deletion/reinstall through game repair
- CPU thermal throttling — not occurring during drops
- GPU thermal throttling — not occurring
- CPU power throttling — not correlated with drops
- CPU/GPU clocks — remain stable
- pagefile — Windows managed
- Windows Memory Compression — was disabled, re-enabled; general smoothness improved but the periodic FPS drop remained
- disk paging / SSD load — not responsible
- background network activity — not correlated
- issue also occurs offline/in Firing Range
Important history
The problem existed previously while using the EA App version of Apex.
After completely reinstalling Apex through Steam, the game initially ran properly and the issue disappeared for a period of time.
Later, the same periodic problem returned again, possibly following an Apex update.
During this period I did not change my NVIDIA driver or Windows version/update configuration, which makes a recent driver or Windows update unlikely to be the trigger.
Expected behavior
Apex should maintain approximately the configured FPS cap without a predictable 10–15 second CPU/memory stall every ~315 seconds.
Actual behavior
Every ~315 seconds, r5apex_dx12.exe enters a very large soft/transition page-fault burst, CPU-side frame time rises, the GPU waits for work, and FPS drops significantly for ~10–15 seconds before recovering.
Request
Please investigate whether Apex currently has a periodic memory-management / asset-streaming / DX12 resource-management task occurring roughly every five minutes that can generate very large transition-page-fault bursts on some systems.
I can provide:
- HWiNFO CSV logs
- Windows PerfMon .blg / CSV logs
- WPR .etl trace
- screenshots from Windows Performance Analyzer
- exact timestamps of reproduced events
The issue is highly repeatable and has been reproduced across multiple sessions.