Latency, throughput, and what to measure

Throughput

Throughput is how many bits transfer per second under the test conditions. It rises with clean signal, available spectrum, and uncongested paths — and falls when any of those weaken.

Download and upload are not symmetric. Many consumer tests emphasize download because media consumption dominates, but uploads matter for video calls, cloud backups, and live reporting.

Latency and jitter

Latency is the round-trip delay to a destination. Jitter is variation in that delay. Real-time calls and games care about both. A high download score with unstable latency can still feel uneven for interactive use.

Geography matters in Chile: distance to an overseas server adds baseline delay that no local radio upgrade can fully erase. Prefer comparing like with like — same server region when possible.

How to test more carefully

• Run several tests at different hours.
• Note whether you are indoors or outdoors.
• Record the radio indicator (for example 4G/5G) without treating it as a speed rating.
• Compare the same test server when possible.
• Repeat after moving a few meters; small location changes can matter indoors.

Interpreting mixed results

A test with high download throughput and elevated latency can still frustrate video calls. Conversely, modest throughput with stable low latency may feel fine for messaging and browsing. Match the metric to the activity you care about.

Airplane mode toggles and forced network selections (when available in device menus) can help you see whether a stubborn attachment to a distant cell is part of the problem — with the caveat that menus differ by handset and software version.

What entei recommends readers remember

Measurement is a tool for learning, not a courtroom verdict about an entire national network. Use tests to understand your context, then read mechanism pages — network layers, speed drivers, and coverage patterns — to interpret the numbers.