How to Reduce Latency for Cloud Gaming

By Paul Snyman · Published August 9, 2026 · 6 min read

Cloud gaming has no local processing to hide behind - every input travels to a remote server and back before you see the result. That round trip is latency, and unlike a locally-run game, there's no buffering it away. Here's what's actually within your control.

What Latency Numbers Mean in Practice

Use Ethernet If You Can

This is the single biggest lever available to most people. WiFi introduces variability - interference, signal degradation with distance, other devices contending for the same airtime - that a wired connection simply doesn't have. Plugging your console, PC, or streaming box directly into the router with an ethernet cable removes an entire category of latency variance, even if your WiFi already feels "fine" for browsing.

Reduce What's Competing for Your Connection

A large download, a backup running in the background, or another device streaming 4K video can all add latency and jitter to your gaming session, even on a fast connection - shared bandwidth means shared contention. Pause anything non-essential before a session, and check LAN Scanner if you suspect an unexpected device is active on the network.

Check the Route, Not Just the Speed

A fast download speed doesn't guarantee low latency - they measure different things. If a specific cloud gaming service consistently feels laggy while your general internet feels fine, MTR traces the route and shows per-hop latency, which can reveal whether one specific hop along the way to that service is adding delay, rather than your connection as a whole.

What's Genuinely Out of Your Control

If you're geographically far from the service's nearest server region, no amount of local optimization fully closes that gap - some cloud gaming platforms let you manually select a server region, which is worth checking if one is available closer to you. Beyond that, congestion on your ISP's network during peak hours is also outside your control, though Speed Test run at different times of day can at least confirm whether that's a contributing factor.

Bufferbloat Is Usually the Real Culprit

If your connection is fast and your gaming still stutters, the most likely explanation is bufferbloat, and it is the one thing on this page that can transform a connection rather than improve it slightly.

Here is the mechanism. Network equipment has memory to hold packets when a link is briefly busy. When the buffers are too large, which is extremely common in consumer routers and ISP modems, a saturated link does not drop packets, it queues them. Your game's packets sit behind whatever else is being transferred, and latency that measured 20 ms idle can become 300 ms or worse while somebody uploads a video.

The signature is unmistakable once you know it. Ping is fine when nothing else is happening and terrible whenever the connection is in use. A speed test that ends and leaves the connection feeling wrong for a moment afterwards is the same effect.

The fix is Smart Queue Management^[source], offered on many routers as SQM, adaptive QoS, or simply bufferbloat control. It works by deliberately holding your throughput slightly below the line's real capacity so the queue never forms at the modem. You give up a small amount of peak speed and get a connection whose latency stays flat under load, which for cloud gaming is a trade worth making every time.

Measure Latency Under Load, Not at Rest

An idle ping is close to meaningless for this. The number that predicts whether a cloud gaming session will be playable is latency while the connection is busy, because that is the state it will be in.

Run a continuous ping and start a large download at the same time. A connection with well-managed queues barely moves. One with bufferbloat climbs immediately and stays high until the transfer ends. That single test tells you more about your gaming experience than any speed test, and it is the before-and-after measurement for any change you make.

Watch jitter as much as the average. Cloud gaming streams video one way and inputs the other, and both are continuous. A steady 45 ms is comfortable. An average of 30 ms that swings between 10 and 90 is not, because the buffer at the far end has to absorb the variation and the way it does that is by adding delay or dropping frames.

Wi-Fi Choices That Matter for Latency

If a cable is genuinely impossible, the Wi-Fi decisions that affect latency are not the same ones that affect throughput.

Use 5 GHz or 6 GHz, always. Not for the headline speed but because 2.4 GHz is congested, and congestion on a shared medium means waiting, and waiting is latency.

Be close to the access point. A weak signal means retransmissions, and every retransmission is a delayed packet rather than a lost one. This is why a session can be smooth in one room and unplayable two rooms away at what looks like a usable signal.

Count what else is on the band. Wi-Fi is a shared medium: only one device transmits at a time on a channel. A household streaming on three televisions is not just using bandwidth, it is taking turns you also need.

Do not use a mesh node over a wireless backhaul if you can avoid it. A node that reaches the router over Wi-Fi doubles every hop's exposure to all of the above. A node with a cable back to the router does not.

Related Articles

• How Many Devices Can Connect to WiFi Before It Slows Down?
• Is It Your WiFi, Your Router, or Your ISP?
• MTR App for iPhone
• PS5 Keeps Disconnecting From WiFi: How to Prove Where It Drops
• Nintendo Switch Keeps Disconnecting From WiFi: Find the Cause