High ping on Windows 11 — elevated latency that makes games feel laggy, video calls choppy, or real-time communication unreliable — is worth diagnosing precisely rather than trying fixes randomly. The difference between 50ms and 200ms ping is significant, and the cause of each is different. This fits into the wider topic we cover in our Complete Guide to Fixing Windows, Browser, and Software Errors.
First: where is the high ping? Open Command Prompt and run:
ping 8.8.8.8Simultaneously run:
ping 192.168.1.1Compare results. If the router ping (192.168.1.1) is under 2ms but the internet ping (8.8.8.8) is high: the problem is outside your network — your ISP or the path to the destination. If the router ping itself is 50ms+: the problem is your local connection. These have completely different fixes.
Local network latency — fix this first
Wi-Fi adds latency. Ethernet ping to the router is typically under 1ms. Wi-Fi ping is typically 2–10ms under good conditions but spikes to 50–100ms+ under interference or in power-saving mode. If you’re on Wi-Fi and this is latency-sensitive (competitive gaming, live video editing), ethernet eliminates the variable immediately.
Wi-Fi power saving specifically adds latency: Control Panel → Power Options → active plan → Change advanced power settings → Wireless Adapter Settings → Power Saving Mode → Maximum Performance. Also Device Manager → Wi-Fi adapter → Properties → Power Management → uncheck “Allow the computer to turn off this device to save power.” After applying both: re-ping and compare.
Network adapter interrupt coalescing
Interrupt Moderation Rate — visible in Device Manager → Network Adapter → Properties → Advanced tab — controls how often the adapter generates CPU interrupts to process received packets. Higher moderation = fewer interrupts = lower CPU usage = higher latency. For latency-sensitive use: set this to “Low” or “Off” (if available). The tradeoff is slightly higher CPU usage. On modern CPUs, this tradeoff is negligible for gaming and video calls.
While in the Advanced tab: also check “Receive Side Scaling” (should be Enabled) and “Energy Efficient Ethernet” (disable this — it reduces adapter activity in ways that spike latency intermittently).
Background applications consuming bandwidth
High ping is often caused by bandwidth saturation rather than a latency problem in the traditional sense. When the upload or download channel is full, new packets have to queue — adding latency. A cloud backup actively uploading, a Windows Update downloading in the background, or another device on the network streaming 4K video all cause this buffering latency.
Task Manager → Performance → Ethernet or Wi-Fi → check the “Send” and “Receive” values during the high ping period. If either is near the connection’s maximum speed: something is saturating the connection. Task Manager → Processes → sort by Network → identify and temporarily close the high-bandwidth processes before gaming or calling.
Buffer bloat — the underdiagnosed cause
Buffer bloat is a specific type of latency caused by oversized network buffers at the router level. When the router’s buffer is full, it holds packets in a queue rather than dropping them — adding hundreds of milliseconds of queue delay that manifests as high ping specifically under load. Idle ping looks fine; under load, ping spikes dramatically.
Test: run a speed test (fast.com) and simultaneously ping during the test. If ping spikes from 20ms to 200ms+ during the download: buffer bloat is present. Fix: enable Smart Queue Management (SQM) or fq_codel on the router if supported (common on routers running DD-WRT, OpenWrt, or higher-end consumer routers). This actively manages the buffer to keep latency low under load.
QoS settings to prioritise your traffic
Router QoS (Quality of Service) can prioritise specific devices or traffic types. If other household members or devices are using bandwidth heavily during your gaming or calls: router admin panel → QoS → add your device to a high-priority queue. This ensures your latency-sensitive traffic isn’t competing equally with bulk downloads on the same connection.
DNS lookup latency
DNS resolution adds latency to every new connection. Slow DNS (common with ISP-provided servers) adds 50–200ms per lookup, making initial connections feel slow even when ongoing ping is low. Switch to a faster resolver:
- Settings → Network and internet → Ethernet or Wi-Fi → DNS server assignment → Manual → set Preferred DNS: 1.1.1.1, Alternate DNS: 8.8.8.8
DNS improvements show up as faster page load times and faster initial game server connections, not necessarily lower continuous ping. Use a tool like DNSBenchmark to verify whether the ISP DNS is actually slow compared to alternatives on your specific network.
Network card and driver settings
Downloading the latest NIC driver from the manufacturer’s support page (rather than Windows Update’s version) sometimes includes network scheduling improvements. Also: administrator Command Prompt →
netsh int tcp show globalConfirm “Receive Window Auto-Tuning Level” is “normal.” If disabled or restricted, run:
netsh int tcp set global autotuninglevel=normalThis ensures Windows optimises the TCP receive window size for the connection’s latency and bandwidth rather than using a fixed (often suboptimal) value.
For the gaming-specific drops and disconnections that accompany high ping under load, our internet drops during gaming guide covers the NIC power management and QoS configuration in more depth. For the router-side configuration including buffer bloat management and DHCP optimisation, our router guide covers the firmware and QoS settings. Microsoft’s Windows 11 networking documentation covers the netsh int tcp settings and their interaction with the Windows networking stack in detail.
Geographic distance to game servers
Physics sets a hard floor on ping: light through fiber cable travels approximately 200km per millisecond of round-trip time. Connecting to a game server 5,000km away has a minimum ping of around 50ms just from physical distance — no local fix reduces this. Many games allow server region selection; choosing a server geographically closer is the only effective solution for distance-based latency.
Also: games that use cloud regions (like AWS, Azure, or Google Cloud) sometimes route to a region that isn’t the geographically closest one. Using a tool like WinMTR to trace the route to the game server shows the network hops — if traffic is being routed through an unexpected geographic path (US traffic routing through Europe before reaching the server), ISP routing issues rather than distance are the cause.
ISP routing inefficiency
ISPs don’t always route traffic optimally. Traffic to a specific destination might take a longer path than necessary due to peering agreements, network topology, or misconfiguration. WinMTR (free) or tracert (built into Windows) shows the path your traffic takes and where latency is added:
tracert 8.8.8.8Each line shows a network hop with its latency. A hop that adds 80ms+ while nearby hops are under 10ms indicates a problem at that hop’s location — either a slow routing node or a long-distance link at that point in the path. This information is useful when reporting to the ISP — specific hop data makes the complaint actionable.
Nagle’s algorithm and real-time applications
Nagle’s algorithm (TCP’s packet coalescing mechanism) delays small packets to combine them into larger, more efficient transmissions. For bulk data transfer, this is efficient. For real-time interactive applications (games, VoIP, remote desktop), the delay adds to latency perceptibly — 5–40ms of additional latency from packet-combining delays.
Disable Nagle’s algorithm for specific applications via registry: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesTcpipParametersInterfaces → find the interface for your network adapter → add DWORD values: “TcpAckFrequency” = 1, “TCPNoDelay” = 1. Restart the network adapter. These settings force immediate packet transmission without buffering. Note: this increases small packet overhead and is only beneficial for latency-sensitive applications — for general browsing, leaving Nagle’s algorithm enabled is correct.
ISP peak hours
Cable internet (not fiber) shares physical infrastructure among neighbourhoods. During peak hours (evenings, weekends), the shared segment fills up and latency increases for everyone on it. High ping specifically in the evening that improves at midnight or early morning confirms this pattern. Cable ISPs monitor congestion data — reporting the specific times and latency values contributes to infrastructure upgrade decisions. Switching to fiber (not shared in the same way) is the permanent solution if peak-hour congestion is consistently problematic.
Windows scheduling and network priority
Windows Game Mode (Settings → Gaming → Game Mode → On) changes the scheduler to prioritise the game’s CPU threads, which indirectly helps network responsiveness during gaming — the game gets more consistent CPU time for its networking operations. MULTIMEDIA CLASS SCHEDULER (MMCSS) manages thread scheduling for media applications including games. Ensure Game Mode is on for gaming sessions.
Also: Network Throttling Index in the registry. HKEY_LOCAL_MACHINESOFTWAREMicrosoftWindows NTCurrentVersionMultimediaSystemProfile → NetworkThrottlingIndex → if set to 10 (default) or a value reducing network priority for games, changing to ffffffff disables network throttling entirely. Restart required. This is a more aggressive setting that improves network responsiveness but may affect background downloads during gaming.
Timing of ping spikes
Ping spikes — ping that’s normally 20ms but periodically jumps to 200ms+ — have different causes than consistently elevated ping. Spikes often correlate with:
- Windows Defender scan starting (CPU scheduling spike)
- Cloud sync uploading (bandwidth spike)
- Background application initiating a connection
- Wireless channel congestion as a neighbour’s router transmits
Monitor ping during a spike: Task Manager → Performance → Disk and Network. If a spike in disk or network usage coincides with a ping spike: that process caused it. Scheduling antivirus scans to off-gaming hours and using ethernet to eliminate wireless channel contention reduces spike frequency significantly.
VPN and ping
VPN always adds latency. The additional latency equals: latency to VPN server + VPN processing overhead + latency from VPN server to destination. A VPN server 200ms away from you adds at least 200ms to every connection through it. For gaming and latency-sensitive applications: disconnect the VPN entirely. No VPN configuration or protocol choice eliminates the distance penalty from routing through a VPN server — it’s physics, not a fixable setting.
If the VPN is required for work: split tunneling (routing only work traffic through the VPN while other traffic goes direct) eliminates the gaming latency penalty while keeping work access through the VPN. Most corporate VPN clients support split tunneling though IT may control whether it’s enabled.
Measuring latency properly
Ping in Windows Command Prompt sends 4 ICMP packets. That’s not enough for meaningful latency analysis. For proper testing: run a longer continuous ping (ping -t 8.8.8.8 — runs until you press Ctrl+C), let it run for at least 2 minutes, look at average, minimum, and maximum values. The range between minimum and maximum (jitter) matters as much as the average — a connection with average 30ms ping but ranging 10–200ms has high jitter that’s worse for gaming than a stable 50ms ping with 45–55ms range.
Tools like PingPlotter continuously monitor ping and display it as a graph over time, making it much easier to identify patterns (spikes at specific times, gradual increases over a session) than Command Prompt’s scrolling output. A 30-minute PingPlotter session during gaming often reveals which specific trigger causes latency increases that felt random.
Malware and ping impact
Malware running on the machine can consume CPU, bandwidth, and network connection slots in ways that increase latency. Cryptocurrency miners particularly increase CPU load that affects the network scheduler’s responsiveness. If ping is consistently high and no obvious cause has been identified: run a Malwarebytes full scan. A clean scan eliminates this cause; finding something explains the latency.
One specific indicator: high ping with high CPU usage but no obvious CPU-heavy applications in Task Manager → a process may be hiding. Autoruns (Sysinternals, free) shows every program that runs at startup, including those that hide from Task Manager. Anything in orange in Autoruns has no Microsoft-verified signature and is worth investigating.
The realistic expectation for what’s achievable with local Windows optimisations: on an ethernet connection, 5–20ms reduction from NIC driver and TCP stack tuning is achievable. On Wi-Fi, switching to Maximum Performance power mode can reduce average latency by 10–30ms and more importantly reduce spike frequency. QoS configuration on the router can reduce the 200ms+ spikes from buffer bloat to under 50ms. But if the baseline ping to the game server is 150ms and the server is genuinely distant: local optimisations bring it to 130–140ms, not 20ms. Some latency is architectural and the only solution is selecting closer servers.
One last practical note: many “high ping” complaints in games are actually high frame time rather than network latency — the game’s frame rendering takes long enough that inputs feel delayed even when network ping is normal. Check the game’s built-in ping display (if available) alongside the Windows ping test. If the game shows 200ms ping but ping 8.8.8.8 shows 20ms: the game’s connection to its specific server is slow, not your general internet. If both show high ping: your internet is the issue. If ping is fine but the game feels laggy: frame time or CPU scheduling is the problem, not network, and the fix is entirely different (GPU driver, game settings, frame rate cap). Our guide on Windows 11 Network Adapter Keeps Disabling covers an adjacent issue.
