The geographical location of the enterprise server will affect the response speed, user experience and operating costs of the enterprise business. The United States and Japan are the key nodes of the Asia-Pacific and European and American markets, and the difference in server performance is easy to lead the enterprise into the decision-making world. The following is a reference of the actual data and network architecture analysis of the US server and the Japanese server, comparing the speed, latency performance and usage scenarios of the two.

1、 the basic differences between geographical location and network architecture

The core differences between US and Japanese servers stem from their geographic location and network infrastructure:

American server: Los Angeles, Silicon Valley, New York as the core, relying on trans-Pacific cable (such as FASTER, TPE) to connect Asia, through the Atlantic cable (such as MAREA) radiation Europe, forming a global network hub. Its advantage is that it covers European and American users with extremely low delay (such as the delay of about 70ms from New York to London), and the total international export bandwidth accounts for more than 40% of the world.

Japanese server: Tokyo, Osaka data center through the Asia-Pacific optical cable ring (such as APG, SJC) directly connected to China, South Korea, Southeast Asia, the physical distance advantage is significant. For example, the delay from Tokyo to Shanghai is about 35ms and to Singapore is about 60ms, making it a natural transit point for regional traffic in Asia.

Key data comparison:

International bandwidth resources: the United States single data center export bandwidth generally more than 500Gbps, Japan is 200300Gbps.

Backbone network density: The United States has more than 30 trans-oceanic cables, and Japan mainly relies on 10 Asia-Pacific regional cables.

2、Comparison of global access delay measurement

Through the ICMP Ping test of 15 test nodes (including China, Europe, the United States, and Southeast Asia), the server latency of the two countries is as follows:

Asian user access delay

China Region:

Beijing to West Coast (Los Angeles) : 120150ms

Beijing to Tokyo: 3550ms

Shanghai to West Coast of USA: 90120ms

Shanghai to Tokyo: 2540ms

Southeast Asia:

Singapore to West Coast of USA: 180220ms

Singapore to Tokyo: 6080ms

Conclusion: The latency advantage of Japanese servers for Asian users is significant, especially in eastern China and Southeast Asia, the speed difference can be more than 3 times.

User access delay in Europe and the United States

European Region:

London to US East Coast (New York) : 7090ms

London to Tokyo: 220260ms

North America:

Los Angeles to New York: 4060ms

Los Angeles to Tokyo: 100120ms

Conclusion: The American server occupies the absolute advantage in the domestic and trans-Atlantic transmission in Europe and the United States, and the delay is only 1/3 to 1/2 of the Japanese server.

Oceania and South America users

Sydney to West Coast: 140170ms

Sydney to Tokyo: 90110ms

Sao Paulo to East Coast: 120150ms

Sao Paulo to Tokyo: 280320ms

Conclusion: Japanese servers are more attractive to Australian users, while the South American market needs to prefer US nodes.

3、 network speed and stability performance

Download speed test (100Mbps bandwidth environment)

Asian users download US server files:

Peak speed: 68 MB/s (subject to international exit congestion)

Stability: During peak hours (18:0022:00 UTC), volatility can reach 30%

Asian users download Japanese server files:

Peak speed: 1012 MB/s

Stability: Volatility is less than 10%

European and American users download Japanese server files:

Peak speed: 34 MB/s

Stability: The trans-Pacific link jitter frequently, retransmission rate of about 5%

Network Jitter and packet loss rate

Us server: Jitter ≤10ms, packet loss rate 0.1%; Asian region jitter 2050ms, packet loss rate 1%3% (cable congestion period)

Japanese server: Asia region: jitter ≤5ms, packet loss rate 0.05%; Europe and the United States: Jitter 3080ms, packet loss rate 2%5%.

The difference in server performance between the United States and Japan symbolizes the imbalance of the global network architecture, and enterprises should determine the distribution of target users when making decisions, if it is mainly the Asian market, the low latency and stability of Japanese servers are higher, if you want to cover the world or focus on European and American users, the cost and bandwidth advantages of American servers are more prominent.