Round-trip time from your browser to every Amazon Web Services region — 40 of them, 13 in Asia, 9 in Europe, 9 in North America, 3 in the Middle East, 3 in Oceania, 2 in South America, 1 in Africa — each against the speed-of-light floor its distance sets, so a slow region is visibly far rather than badly routed.
A browser hands JavaScript no raw socket, and an ICMP echo needs one — so no web page can send a true ping, to AWS or anywhere. What gets timed instead is a small HTTPS request to DynamoDB’s regional endpoint. A real service answers it, which costs a little more than an echo would and is the point: it is closer to what an API call from here will actually feel like.
dynamodb.{region}.amazonaws.com answers 200 with 'healthy: dynamodb.{region}.amazonaws.com' and Access-Control-Allow-Origin: *, so the reply names the region that served it.
The first request of a sweep is discarded — it pays for DNS, TCP and TLS — so every sample after it is essentially one round trip, and the reported figure is the fastest of them.
| Region | City | Country | IPv4 prefixes |
|---|---|---|---|
| af-south-1 | Cape Town | ZA | 140 |
| ap-southeast-7 | Bangkok | TH | 72 |
| cn-north-1 | Beijing | CN | 132 |
| ap-east-1 | Hong Kong | HK | 161 |
| ap-south-2 | Hyderabad | IN | 102 |
| ap-southeast-3 | Jakarta | ID | 119 |
| ap-southeast-5 | Kuala Lumpur | MY | 91 |
| ap-south-1 | Mumbai | IN | 303 |
| ap-northeast-3 | Osaka | JP | 164 |
| ap-northeast-2 | Seoul | KR | 266 |
| ap-southeast-1 | Singapore | SG | 415 |
| ap-east-2 | Taipei | TW | 74 |
| ap-northeast-1 | Tokyo | JP | 428 |
| cn-northwest-1 | Yinchuan, Ningxia | CN | 124 |
| eusc-de-east-1 | Brandenburg an der Havel | DE | 66 |
| eu-west-1 | Dublin | IE | 533 |
| eu-central-1 | Frankfurt | DE | 578 |
| eu-west-2 | London | GB | 406 |
| eu-south-1 | Milan | IT | 222 |
| eu-west-3 | Paris | FR | 216 |
| eu-north-1 | Stockholm | SE | 213 |
| eu-south-2 | Zaragoza (Aragon) | ES | 148 |
| eu-central-2 | Zurich | CH | 85 |
| me-central-1 | Dubai | AE | 85 |
| me-south-1 | Manama | BH | 171 |
| il-central-1 | Tel Aviv | IL | 94 |
| us-east-1 | Ashburn, Virginia | US | 1,422 |
| us-gov-east-1 | Ashburn, Virginia | US | 102 |
| ca-west-1 | Calgary, Alberta | CA | 78 |
| us-east-2 | Columbus, Ohio | US | 608 |
| ca-central-1 | Montreal, Quebec | CA | 200 |
| us-gov-west-1 | Portland, Oregon | US | 133 |
| us-west-2 | Portland, Oregon | US | 663 |
| mx-central-1 | Queretaro | MX | 72 |
| us-west-1 | San Jose, California | US | 571 |
| ap-southeast-6 | Auckland | NZ | 68 |
| ap-southeast-4 | Melbourne | AU | 96 |
| ap-southeast-2 | Sydney | AU | 333 |
| sa-west-1 | Santiago | CL | 35 |
| sa-east-1 | Sao Paulo | BR | 299 |
Prefix counts come from AWS’s own published feed. Each region links to its distance, its physical floor and the endpoint used to time it.
Open the instrument, select AWS, choose regions and run a sweep. Nothing is pre-selected and nothing runs until you press the button.