Guides
Download vs Upload Speed
Understand download and upload, estimate transfer times and investigate differences between your plan, speed test and apps.
Download moves data from the network to your device; upload moves data from your device to the network. The two speeds describe different directions, so a fast download does not guarantee a fast upload. Start with the task that is struggling: are you receiving data, sending it, or doing both?
Which direction does your task use?
- Watching video, opening websites and downloading games or files mainly use download. Upload is still relevant: TCP also sends acknowledgements back for received data.
- Backing up to the cloud, sending files and broadcasting live video mainly use upload. Retrieving a cloud backup or file uses download. A sync app can transfer data in both directions.
- A video call sends your audio and video and receives other participants' streams. Both directions, delay and stability can therefore matter. Requirements depend on the app and chosen quality; one speed target does not cover every call.
Speed and file size use different units
Mbps means millions of bits per second; MB means millions of bytes. One byte contains eight bits. A 1 GB file, where GB means 1,000,000,000 bytes, contains 8,000 megabits. At a constant 100 Mbps, transferring the data alone ideally takes 80 seconds; at 10 Mbps it takes 800 seconds, or 13 minutes and 20 seconds. These are calculations, not promises: actual throughput, the service, your device and processing affect completion time. MiB and GiB are different size units; do not substitute them for MB and GB.
Read both speeds in your actual plan
- Compare the stated download and upload separately, using the same units. A hypothetical 100/10 Mbps plan has different rates; 100/100 Mbps has equal stated rates. These examples are not typical speeds for a particular technology.
- Different download and upload rates do not prove a fault. Labels such as fibre, cable, DSL or 5G do not by themselves establish the two rates your plan offers or the speed your device can measure.
- One measurement does not independently establish whether a contract is being met. Compare repeated, documented results with the provider's definitions, conditions and measurement method before concluding that delivery falls short of the agreement.
Separate the speed test from an app transfer
VMAXCHECK uses M-Lab NDT to transfer data over one TCP connection to a test server at the time of measurement. It does not measure every app's server, the household's total simultaneous capacity or the cause of a fault. Record the time, server, device and whether you used Wi-Fi or Ethernet. A missing upload result is not zero. The test creates traffic itself, so run it outside an active call, video session or game. M-Lab publishes results and the public IP address used. In an app, distinguish actual transfer from waiting, file verification and installation.
Narrow the problem down one change at a time
- Repeat the test on the same device, with the same method and connection type, without your own background transfers. Record the server and time and compare repeats under similar conditions. A low upload reading or one difference does not locate a fault.
- Briefly pause your own backup or sync and try the affected app again without a simultaneous speed test. A busy upload can build queues and affect other traffic, including video playback. An improvement points towards load worth investigating; it does not prove a particular router defect.
- Compare Wi-Fi and Ethernet on the same device. Check the active connection and the adapter and port capacities. A cable into a mesh unit can still leave a wireless link further along the path. Use the comparison to narrow the issue before buying equipment.
- If only one service is slow, try it on another suitable device and try another service on the original device. Check the app's own speed limits and service status. OneDrive, for example, can cap its sync rate; work or school settings may be controlled by IT. Keep organisational rules in place and record settings before any permitted changes.
Choose support and a plan around the task
Keep comparable download and upload readings, the time and connection, and the app's specific error. If both directions already support the task, a more expensive plan is not automatically the answer. If repeated comparisons show problems across several services, contact the provider with your results; if they affect only one app or device, involve its support. Do not share passwords or private files. A good speed test does not prove that every service works well.