Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteiTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
To reduce bufferbloat, first measure latency when the connection is idle and while it is busy. Then enable Smart Queue Management (SQM)—preferably CAKE or fq_codel—if your router supports it, and set upload and download shaper rates slightly below the real bottleneck speeds. Retest and adjust until loaded latency is controlled without sacrificing useful throughput.
SQM is different from traditional device or application-priority QoS. It controls the queue itself, so you do not need a long list of gaming, streaming, or device rules. If your router cannot run SQM, verify whether its firmware can be upgraded or replace it with hardware that can sustain your internet speed with SQM enabled.
What bufferbloat is and why it causes lag
Bufferbloat is excess latency created when a router, modem, access point, or other network device stores too many packets in a queue. The problem appears when traffic arrives at a bottleneck faster than the bottleneck can transmit it. Instead of being discarded immediately, packets wait in a long first-in, first-out queue. A web request, voice packet, or game update can therefore sit behind a large upload or download even though the connection has plenty of nominal bandwidth.
Bufferbloat can occur in either direction. A full upload queue may delay acknowledgements, video calls, and game traffic, while a download queue can delay packets arriving from the internet. Wi-Fi introduces another possible queue: airtime and per-station transmit queues can become congested even when the ISP-facing queue is well managed.
#1 Best Overall
- OneMesh Compatible Router - Form a seamless WiFi when work with TP-Link OneMesh WiFi Extenders
- Next-Gen Wi-Fi 6 Technology – The Archer AX10 leverages advanced Wi-Fi 6 features like OFDMA and 1024-QAM to deliver improved efficiency across your entire network. Perfect for high-bandwidth activities like streaming, gaming, and smart home connectivity.
- Next-gen Dual Band router - 300 Mbps on 2. 4 GHz (802. 11n) plus 1201 Mbps on 5 GHz (802. 11ax)
- Connect more devices than ever before - Wi-Fi 6 technology simultaneously communicates more data to more devices using OFDMA and MU-MIMO while reducing lag dramatically
- Powerful Dual-Core 900MHz Processor – Handles multiple data streams simultaneously for reliable performance across your devices. Ensures smooth streaming, online gaming, and video conferencing without buffering or lag.
The useful measurement is not speed alone. You need to compare latency at idle with latency while uploading and downloading. A connection that delivers its advertised rate but adds hundreds of milliseconds under load can feel much worse than a slower connection with a short, well-managed queue.
Why SQM is not the same as traditional QoS
Traditional QoS usually means priority rules such as “put this console first” or “prioritize voice over downloads.” Those rules can move selected packets ahead of others, but they do not necessarily stop a sender from filling a very long queue. The queue still exists; priority traffic is simply allowed to pass through it sooner.
Smart Queue Management takes a different approach. It deliberately controls the bottleneck rate and divides traffic into managed per-flow queues. CAKE and fq_codel can schedule those queues so that one large transfer is less able to make every other flow wait. The goal is to create a small, controllable queue inside your router rather than allowing an uncontrolled queue to form in upstream hardware.
Recommended Free Tools
“The goal is to set the software shaper to a bandwidth that is slightly lower than the actual (bloated) bottleneck in the hardware, so we can control the queue using FQ-CoDel or CAKE.”
Bufferbloat.net SQM tuning guidance
In this article, “without router QoS” means without conventional device and application-priority rules. SQM is still a traffic-management function, but it solves the queue problem by shaping and scheduling traffic rather than by maintaining a large priority list.
Rank #2
- DUAL-BAND WIFI 6 ROUTER: Wi-Fi 6(802.11ax) technology achieves faster speeds, greater capacity and reduced network congestion compared to the previous gen. All WiFi routers require a separate modem. Dual-Band WiFi routers do not support the 6 GHz band.
- AX1800: Enjoy smoother and more stable streaming, gaming, downloading with 1.8 Gbps total bandwidth (up to 1200 Mbps on 5 GHz and up to 574 Mbps on 2.4 GHz). Performance varies by conditions, distance to devices, and obstacles such as walls.
- CONNECT MORE DEVICES: Wi-Fi 6 technology communicates more data to more devices simultaneously using revolutionary OFDMA technology
- EXTENSIVE COVERAGE: Achieve the strong, reliable WiFi coverage with Archer AX1800 as it focuses signal strength to your devices far away using Beamforming technology, 4 high-gain antennas and an advanced front-end module (FEM) chipset
- OUR CYBERSECURITY COMMITMENT: TP-Link is a signatory of the U.S. Cybersecurity and Infrastructure Security Agency’s (CISA) Secure-by-Design pledge. This device is designed, built, and maintained, with advanced security as a core requirement.
Step 1: Establish a reliable bufferbloat baseline
Measure before changing settings. Use a test that reports latency when idle and during both download and upload. Bufferbloat.net lists Waveform, Speedtest.net, Cloudflare Speed Test, and LibreQoS testing options. The exact labels differ, but each test should let you see whether latency rises under load.
- Use Ethernet if possible. Connect a computer directly to the router with a wired cable. This minimizes Wi-Fi airtime and driver effects, helping you isolate the ISP or router queue.
- Stop unrelated traffic. Pause cloud backups, software updates, video streams, torrents, and other uploads or downloads on every device. If you cannot stop them, record that fact because it makes before-and-after comparisons less reliable.
- Record idle latency. Run the test with no deliberate transfer and note the idle latency, download latency, upload latency, and measured throughput.
- Run the loaded test. Let the test saturate download, upload, or both. Record the highest or sustained loaded latency rather than looking only at the final speed score.
- Repeat under the same conditions. Use the same computer, cable, test service, time of day when practical, and router configuration for every retest.
Bufferbloat.net’s 2026 guidance treats loaded latency below roughly 15–25 milliseconds as an indication that bufferbloat is under control. That is a source-specific rule of thumb, not a universal guarantee: access technology, geographic distance, Wi-Fi conditions, and the test server can all add latency that SQM cannot remove.
Free tools Windows power users keep installed
One-click scans. No signup required.
| Record | Why it matters |
|---|---|
| Idle latency | Provides the baseline path delay before queues build. |
| Download-loaded latency | Shows whether the downstream bottleneck creates a queue. |
| Upload-loaded latency | Shows whether the upstream bottleneck creates a queue. |
| Idle and loaded throughput | Lets you balance latency control against usable speed. |
| Connection type and test method | Provides context for later comparisons, especially when Wi-Fi or variable-rate service is involved. |
Step 2: Check whether the current router supports SQM
Open the router’s current firmware interface and search for Smart Queue Management, SQM, CAKE, fq_codel, or PIE. Menus vary by firmware and model, so do not assume that a generic “QoS” switch provides the same function. A page containing only device-priority lists may not include a real queue-management shaper.
OpenWrt
On OpenWrt, Bufferbloat.net points to the luci-app-sqm package. After installing that package, the LuCI interface exposes SQM settings where you can select the interface, enter upload and download rates, and choose a queue discipline. The exact interface name depends on your network layout; the WAN device is normally the one connected toward the internet.
Choosing a queue discipline
| Option | Role | Practical implication |
|---|---|---|
| CAKE | Queue management and flow scheduling with extensive traffic-handling features. | A strong default when the router has enough processing capacity and the firmware supports it. |
| fq_codel | Fair-queueing combined with controlled-delay active queue management. | A well-established alternative when CAKE is unavailable or too demanding for the hardware. |
| PIE | Another active queue-management method available on some firmware. | Use only when it is the supported queue-management choice on your platform; compare results by retesting. |
Do not enable every traffic feature at once. Start with one SQM instance on the actual bottleneck interface, then measure. Extra shaping layers—such as an SQM instance on both the main router and a downstream access point—can reduce throughput or make diagnosis harder.
Rank #3
- Next-Gen Gigabit Wi-Fi 6 Speeds: 2402 Mbps on 5 GHz and 574 Mbps on 2.4 GHz bands ensure smoother streaming and faster downloads; support VPN server and VPN client¹
- A More Responsive Experience: Enjoy smooth gaming, video streaming, and live feeds simultaneously. OFDMA makes your Wi-Fi stronger by allowing multiple clients to share one band at the same time, cutting latency and jitter.²
- Expanded Wi-Fi Coverage: 4 high-gain external antennas and Beamforming technology combine to extend strong, reliable, Wi-Fi throughout your home.
- Improved Battery Life: Target Wake Time helps your devices to communicate efficiently while consuming less power.
- Improved Cooling Design: No heat ups, no throttles. A larger heat sink and redefined case design cools the WiFi 6 system and enables your network to stay at top speeds in more versatile environments.
Step 3: Set upload and download shaper rates
Measure the connection with SQM disabled first. The measured rate is a starting point, not a value to copy blindly into the shaper. The shaper must be slightly slower than the real bottleneck so that packets queue in the router, where CAKE or fq_codel can manage them, instead of in the modem or ISP equipment.
- Start below the measured rate. Enter upload and download values modestly below the corresponding measured bottleneck speeds. There is no single percentage that is correct for every cable, fiber, DSL, LTE, or satellite connection.
- Enable both directions when appropriate. A download-only configuration cannot control an upstream queue, and an upload-only configuration cannot control a downstream queue. Configure the directions that your router and firmware expose.
- Run the same loaded tests. Compare download-loaded latency, upload-loaded latency, and throughput with the baseline.
- Lower the relevant rate if latency remains high. If upload latency spikes only during upload, reduce the upload shaper first. If the spike occurs during download, adjust the download value.
- Raise rates cautiously when latency is controlled. Increase a rate in small increments and retest. Stop when loaded latency begins to worsen, then return to the last stable setting.
The ideal rate can change with time, protocol overhead, congestion, and the access network’s scheduling. Bufferbloat.net therefore describes tuning as iterative trial and error: there is often no independent ground truth for the perfect number. Optimize for a useful balance of latency and throughput rather than for the largest speed-test result.
| Result after enabling SQM | Likely adjustment |
|---|---|
| Loaded latency falls substantially but throughput is much lower than needed | Increase the affected shaper rate gradually and retest. |
| Throughput is high but loaded latency still spikes | Lower the shaper rate for the direction showing the spike. |
| Upload is good, download is poor | Tune download separately; do not change upload unnecessarily. |
| Download is good, upload is poor | Tune upload separately; upstream queues are often the limiting factor. |
| Both directions remain poor at every tested rate | Check that SQM is attached to the correct WAN interface and investigate modem, ISP, or Wi-Fi queues. |
What a successful result looks like
A successful configuration does not necessarily produce the highest possible speed score. It keeps loaded latency close to the idle baseline while preserving enough throughput for the household’s normal use. Test interactive traffic—such as a voice call or game—while a large transfer is running if you want to verify the experience directly, but keep the same test conditions when comparing settings.
Do not generalize the historical figures sometimes quoted in SQM discussions. In one Comcast cable setup, Dave Täht’s account on Bufferbloat.net reports a baseline latency of 65 milliseconds, an increase of 10 milliseconds after tuning, and a previous increase of about 1.2 seconds. The page characterizes that particular result as a 110-times improvement for that test setup. It is a case study, not a promise of the same result on your connection.
Fixed-rate connections versus changing-rate connections
Standard CAKE configurations use fixed upload and download rates. That is straightforward when the usable bottleneck speed is stable, as it often is on a well-behaved wired service. It becomes difficult when the link’s capacity changes substantially from minute to minute.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Rank #4
- NIGHTHAWK WIFI 6 ROUTER FOR YOUR WHOLE HOME: Delivers fast, reliable WiFi across every room of your apartment or small home for streaming, gaming, video calls, and smart home devices, all running at the same time without slowing each other down.
- WORKS WITH YOUR EXISTING INTERNET SERVICE: Pairs with your existing modem or gateway via ethernet. Compatible with most cable, fiber, DSL, and satellite providers. Some gateways and modem router combos may require bridge mode. No coax needed.
- SET UP AND MANAGE YOUR NETWORK WITH THE NIGHTHAWK APP: Download the free Nighthawk app on iOS or Android for guided setup. Manage WiFi, run speed tests, pause devices, and set up guest networks from anywhere. Active internet required.
- READY FOR THE DEVICES YOU ALREADY OWN: Your phones, laptops, and TVs work right out of the box. WiFi 6 delivers speeds up to 1.8 Gbps across 2.4 GHz and 5 GHz bands. Backward compatible with WiFi 5 and earlier.
- COVERAGE IN EVERY ROOM: Covers up to 1,500 sq. ft. for up to 20 connected devices. Walls, floors, and interference can reduce range. Larger or multi-story homes may benefit from a NETGEAR Orbi mesh WiFi system.
LTE, Starlink, and some cable services can vary with radio conditions, network load, signal quality, or provider scheduling. A fixed rate that is safe at the worst moment may waste capacity at the best moment; a rate that is fast enough at the best moment may allow queues to build when capacity falls.
The cake-autorate project adjusts shaping parameters using latency measurements. Its documentation describes support for OpenWrt and Asuswrt-Merlin and identifies stable version 3.2.1 on the v3.2 branch. Check the project’s current release information and installation requirements before deploying it, because versions and firmware compatibility can change.
| Connection behavior | Starting approach |
|---|---|
| Speed is usually stable | Use fixed CAKE or fq_codel rates and tune them with repeatable tests. |
| Speed changes frequently | Consider adaptive control such as cake-autorate if your firmware and hardware support it. |
| Speed varies only at busy times | Use conservative fixed rates first, then compare performance at quiet and busy periods. |
Separate ISP bufferbloat from Wi-Fi delay
Run the baseline and retests over Ethernet first. If wired tests show controlled latency but wireless tests still spike, the ISP-facing SQM is not necessarily the problem. Wi-Fi uses shared airtime, and the access point or client driver may maintain per-station transmit queues. Airtime fairness, airtime queue length, interference, channel contention, and client power-saving behavior can all affect latency.
Test one wireless device at a time near the access point and then at the normal problem location. Compare the result with the wired result under the same internet load. If only Wi-Fi is affected, investigate the access point’s wireless queueing and radio configuration rather than continually lowering the WAN shaper rate.
- Check whether latency rises when another wireless client transmits locally, even if the internet is idle.
- Compare 2.4 GHz and 5 GHz or 6 GHz where available, because congestion and range differ.
- Update access-point and client firmware before drawing conclusions from one test.
- Remember that WAN SQM cannot directly remove every delay created inside a wireless network.
If the current router has no SQM
If the firmware offers only basic priority rules, those rules may not solve a long bottleneck queue. First check whether an official or compatible firmware update adds SQM. For OpenWrt, verify that the exact router model, hardware revision, and wireless drivers are supported before changing firmware.
Best Value
- 𝐆𝐢𝐠𝐚𝐛𝐢𝐭 𝐖𝐢𝐅𝐢 𝐟𝐨𝐫 𝟖𝐊 𝐒𝐭𝐫𝐞𝐚𝐦𝐢𝐧𝐠 – Up to 5400 Mbps WiFi for faster browsing, streaming, gaming and downloading, all at the same time. Performance varies by conditions, distance to devices, & obstacles such as walls.
- 𝐅𝐮𝐥𝐥 𝐅𝐞𝐚𝐭𝐮𝐫𝐞𝐝 𝐖𝐢𝐅𝐢 𝟔 𝐑𝐨𝐮𝐭𝐞𝐫 – Equipped with 4T4R and HE160 technologies on the 5 GHz band to enable max 4.8 Gbps ultra-fast connections.Power:12 V 2.5 A
- 𝐂𝐨𝐧𝐧𝐞𝐜𝐭 𝐌𝐨𝐫𝐞 𝐃𝐞𝐯𝐢𝐜𝐞𝐬 – Supports MU-MIMO and OFDMA to reduce congestion and 4X the average throughput
- 𝐄𝐱𝐭𝐞𝐧𝐬𝐢𝐯𝐞 𝐂𝐨𝐯𝐞𝐫𝐚𝐠𝐞 - Covers up to 2,000 sq. ft. High-Power FEM, 6× Antennas, Beamforming, and 4T4R structures combine to adapt WiFi coverage to perfectly fit your home and concentrate signal strength towards your devices.
- 𝐌𝐨𝐫𝐞 𝐕𝐞𝐧𝐭𝐬, 𝐋𝐞𝐬𝐬 𝐇𝐞𝐚𝐭 – Improved vented areas help unleash the full power of the router
If replacing hardware is necessary, choose by verified SQM capability and measured throughput rather than by a generic “gaming router” label. The device must be able to run CAKE or fq_codel at your internet speed with the features you need enabled. Bufferbloat.net identifies GL.iNet routers as a category running modified OpenWrt that includes SQM, but the available evidence does not establish a universally suitable model, current stock, price, or throughput ceiling.
| Option | Advantages | Risks or checks |
|---|---|---|
| Keep current firmware and enable SQM | No hardware purchase and the quickest test. | Confirm the feature is genuine SQM and that the processor can sustain your speed. |
| Install compatible OpenWrt firmware | Can add luci-app-sqm and provide CAKE or fq_codel. |
Verify exact model and revision support; firmware changes can affect Wi-Fi and warranty. |
| Replace the router | Can provide sufficient CPU capacity and current SQM support. | Check exact firmware, WAN speed ceiling, modem compatibility, and return policy before buying. |
Troubleshooting persistent bufferbloat
Latency never improves
- Confirm the SQM service is running and attached to the WAN interface that actually carries internet traffic.
- Verify that the entered units match the firmware’s requirement, such as megabits per second rather than megabytes per second.
- Retest with one wired computer and all other traffic paused.
- Lower the shaper rate enough to move the queue from upstream equipment into the router, then raise it gradually.
Speed collapses after SQM is enabled
- Check whether the router’s CPU is saturated while CAKE or fq_codel is active.
- Use fq_codel instead of CAKE if the hardware cannot process CAKE at the required rate.
- Check whether encryption, traffic inspection, or another acceleration-disabling feature is running at the same time.
- Raise the shaper cautiously only after loaded latency remains acceptable.
Only upload tests fail
Focus on the upstream shaper. Cable and other asymmetric services can have a much smaller upload bottleneck than download, so an upload value copied from the advertised download plan is inappropriate. Re-measure upload independently and tune that direction without changing the working download value.
Only download tests fail
Check the downstream rate and whether the modem or provider equipment is the actual queue. If wired SQM works at low rates but fails near the service maximum, the router may lack enough processing capacity for the required configuration.
Wired tests pass but Wi-Fi tests fail
Treat this as a wireless queue or airtime investigation. Compare clients, bands, distance, and local traffic. A WAN-side shaper may be functioning correctly while a wireless transmit queue remains too long.
A repeatable maintenance checklist
- Save the router configuration before changing SQM settings.
- Record idle latency, loaded upload latency, loaded download latency, and throughput over Ethernet.
- Identify whether the connection’s usable rate is stable or variable.
- Enable one SQM configuration on the correct WAN interface.
- Select CAKE or fq_codel, depending on firmware support and router capacity.
- Enter upload and download rates slightly below measured bottleneck rates.
- Retest under the same conditions and adjust one direction at a time.
- Verify that ordinary browsing, calls, gaming, and large transfers coexist acceptably.
- Repeat a wireless-only test to detect delays that WAN SQM cannot address.
- Recheck settings after major firmware, modem, service-plan, or network-topology changes.
There is no population-wide percentage improvement that can be responsibly promised for SQM. Results depend on the access technology, the true bottleneck, router capacity, Wi-Fi conditions, and how accurately the shaper rate is tuned. The reliable method is measurement, controlled shaping, and repeated verification.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
