Edge Rendering and the Modern Performance Stack: How Today’s Websites Load in Under a Second

A decade ago, making a website faster usually meant buying a bigger server. Today, the fastest websites are built very differently. They deliver content from locations close to each visitor, cache aggressively and intelligently, send images in modern formats sized for each device, and ship only the code a page actually needs.

This combination of techniques — often described as the modern performance stack — makes it realistic for business websites to load meaningful content in under a second for most visitors. Understanding how the pieces fit together helps businesses make better decisions about technology, hosting and development partners.

Why distance still matters

Every request a browser makes travels physically across networks. A visitor in Vancouver requesting a page from a server in Mumbai, or a visitor in Manchester reaching a server in Texas, adds round-trip delay before any content arrives — and a typical page involves many requests. No amount of server power removes the speed of light. The solution is to bring content closer to the user.

Layer one: content delivery at the edge

A content delivery network (CDN) stores copies of a website’s files in data centres around the world, known as edge locations. Visitors are automatically served from the nearest one. For static assets such as images, stylesheets and scripts, this has long been standard practice.

The modern shift is that entire pages — not just assets — are increasingly delivered from the edge. Pages that do not change for every visitor can be pre-rendered and cached globally, so the first byte of a page can arrive in a fraction of the time an origin server would need.

Layer two: edge rendering and smart caching

Not every page can be identical for every visitor. Edge rendering allows lightweight code to run at those distributed locations, personalising or assembling pages close to the user rather than at a single central server. Modern frameworks combine several strategies:

• Static generation for pages that rarely change, such as service pages and articles.
• Incremental regeneration, where cached pages are refreshed in the background when content changes, so visitors never wait for a rebuild.
• Edge or server rendering for genuinely dynamic content such as search results, account areas or real-time pricing.
• Stale-while-revalidate caching, which serves a fast cached response immediately while fetching a fresh version for the next visitor.

The fastest request is the one that never has to travel back to your origin server.

Layer three: modern media

Images are frequently the heaviest part of a page. Modern formats such as WebP and AVIF can deliver the same visual quality at a fraction of the size of older JPEG and PNG files. Responsive images ensure a phone downloads a smaller version than a large desktop display, and lazy loading defers images that are not yet visible. For many websites, optimising media alone produces the single largest improvement in load time.

Layer four: lean, prioritised code

Large JavaScript bundles delay how quickly a page becomes interactive, particularly on mid-range mobile devices. Modern development practices reduce this cost through code splitting, server-side rendering of components so less code reaches the browser, deferring non-essential scripts, and carefully auditing third-party tags for analytics, chat and advertising. Critical styles can be delivered inline so the first view renders immediately, with fonts loaded in a way that avoids invisible text and layout shifts.

Layer five: modern network protocols

HTTP/2 and HTTP/3 allow many resources to load in parallel over a single connection, and HTTP/3 — built on the QUIC protocol — performs better on unreliable mobile networks. Resource hints such as preconnect and preload tell the browser what it will need before it discovers it, and speculative prefetching can load the next likely page before a visitor clicks.

Measuring what matters

Performance work should be guided by Core Web Vitals, measured from real users rather than laboratory tests alone:

• Largest Contentful Paint (LCP) — how quickly the main content appears; good is 2.5 seconds or less.
• Interaction to Next Paint (INP) — how quickly the page responds to taps and clicks; good is 200 milliseconds or less.
• Cumulative Layout Shift (CLS) — how stable the layout is while loading; good is 0.1 or less.

Field data from Google Search Console and real-user monitoring shows how actual customers experience the site across devices and networks, which matters far more than a single score on a fast office connection.

Applying the stack to your website

The good news is that these techniques are not exclusive to large technology companies. A well-architected website can adopt them progressively:

1. Put a CDN in front of the site and enable full-page caching for pages that are the same for every visitor.
2. Convert and resize images into modern formats, and serve them responsively.
3. Audit and reduce JavaScript, removing unused plugins and tags and deferring what remains.
4. Enable HTTP/3 and resource hints through your hosting or CDN provider.
5. Choose a rendering strategy per page type, using static generation wherever content allows.
6. Monitor real-user metrics continuously, and treat regressions as defects.

Speed as a competitive advantage

Customers rarely notice a fast website, but they always notice a slow one. As more competitors invest in modern performance engineering, speed is becoming a baseline expectation rather than a differentiator. Businesses that build on the modern performance stack deliver experiences that feel instant — and that consistently convert, rank and retain better as a result.

VNForge Labs engineers fast, globally delivered websites and improves the performance of existing platforms. Ask us for a performance review.