Latency
Definition
The delay between sending a request and receiving a response. Like the time it takes between asking a question and getting an answer.
Use Cases
- Netflix: Reducing video start time and buffering for global streaming users — Uses a CDN approach (Netflix Open Connect) by placing caching servers close to ISPs and end users to reduce round-trip time and speed up content delivery. (Lower startup delay and fewer rebuffering events, improving viewer experience and reducing transit bandwidth costs.)
- Cloudflare: Improving website and API responsiveness for customers worldwide — Runs a global edge network that terminates TLS and serves cached/static content near users, reducing the distance requests travel and lowering round-trip latency. (Faster page loads and snappier API responses, which can improve conversion rates and reduce origin infrastructure load.)
- Riot Games: Keeping competitive online gameplay responsive and fair — Deploys game servers in multiple regions and uses network peering/optimized routing so players connect to nearby servers, minimizing ping and jitter. (Lower player-reported lag and improved competitive integrity, supporting better retention and esports-grade play.)
Frequently Asked Questions
- What's the difference between latency and bandwidth?
- Latency is the time it takes for data to travel from sender to receiver (often measured in milliseconds). Bandwidth is how much data can be transferred per second (like Mbps or Gbps). Low latency makes interactions feel instant; high bandwidth helps move large files faster. You can have high bandwidth but still feel “lag” if latency is high.
- When should I optimize for low latency?
- Optimize for low latency when users need fast, interactive responses: online gaming, video calls, live trading, real-time dashboards, voice assistants, remote desktops, and APIs that power user interfaces. If your workload is batch processing (nightly jobs, backups), latency usually matters less than throughput and cost.
- How much does latency cost?
- Latency itself isn’t billed, but reducing it can change costs. Common cost drivers include using more regions (extra infrastructure), CDNs/edge services (request and data transfer fees), premium networking or dedicated links, and additional caching layers. Also note that data egress and cross-region traffic can increase costs when you move data closer to users.
Category: networking
Difficulty: basic
See Also