Data visualization is a language of its own — spoken by analysts, engineers, and everyone who works with numbers. For a long time I put off studying charts seriously, assuming they were "a designer's thing." But when I saw the Edge Requests dashboard for my site devitpro.uz on Vercel — with real numbers across 19 regions of the world — I realized: without the ability to read and build charts, any developer remains half-blind. The data is there, but it stays silent until you turn it into a visual image.

In this post, I share my personal experience studying eight types of 3D visualization: from simple volumetric bar charts to normalized stacked graphics. And I do it not with abstract examples, but with real Edge Requests data from devitpro.uz covering June 28 to September 1, 2026. Every chart is built on ECharts and animated with GSAP — the best stack available today for browser-based 3D visualization.

My conviction: a chart is not decoration for a report — it is a thinking tool. A good chart answers a question faster than a table. A bad one lies even when the numbers are correct. The difference lies in the chart type and how it is built.

The Data We Will Work With

Before building charts, it's essential to understand the data. This is real Edge Requests statistics from my site devitpro.uz hosted on Vercel. The dataset covers three months and is distributed across 19 regions of the world. Total for the period — 46,607 requests.

Period Jun 28, 2026 — Sep 1, 2026 (three calendar months)
Regions 19 points of presence across the global Vercel Edge Network
Total Requests 46,607 Edge Requests across the entire observation period
Top Region Hong Kong (hkg1) — 29,043 requests (56.1%)

This data is excellent learning material: it's heterogeneous (a clear leader plus a long tail), covers multiple continents, and lets us showcase every visualization type in action. That's exactly why I chose it for this publication.

Eight Types of 3D Visualization — A Breakdown

Below are eight charts, each answering a different question. I built them all on the same dataset so you can compare how perception shifts depending on the visualization type.

1. 3D Chart — Edge Requests by Month
Volumetric bars display three monthly periods. Bar height = number of requests.
Jun 1 – Jul 1, 2026 — 14,443 · Jul 1 – Aug 1, 2026 — 8,897 · Aug 1 – Sep 1, 2026 — 23,267
2. 3D Diagram — Distribution by Region
A volumetric pie diagram shows each region's share of total traffic.
Top 5: Hong Kong (56.1%) · San Francisco (14.9%) · Washington (7.8%) · Frankfurt (6.8%) · Cleveland (5.1%)
3. Grouped Histogram by Region
Grouped bars let you compare regions side by side by absolute values.
Hong Kong dramatically outpaces every other region — immediately visible.
4. Stacked Histogram
Stacked bars show each region's contribution to the total request volume.
Total bar height = the sum of all requests across the period.
5. Normalized Stacked Histogram
All bars are normalized to 100% — showing percentage share by region.
Ideal for comparing traffic structure across different periods.
6. Volumetric Grouped Histogram
3D bars grouped by region — adds perceptual depth.
The 3D effect helps separate bars with close values.
7. Volumetric Stacked Histogram
Stacking in 3D: you can see both the sum and each region's contribution.
A strong choice for presentations where visual expressiveness matters.
8. Volumetric Normalized Stacked Histogram
3D + stacking + normalization — maximum information in a single chart.
Each bar totals 100%, and the segments show each region's share.

Regional Map of Edge Requests

I've pulled the regional statistics into a separate block — they deserve their own section. Each card shows the number of requests, its percentage of the total. The top 3 regions are highlighted with a golden accent.

Hong Kong hkg1
29,043 56.1%
San Francisco, USA sfo1
7,739 14.9%
Washington, D.C., USA iad1
4,045 7.8%
Frankfurt, Germany fra1
3,543 6.8%
Cleveland, USA cle1
2,649 5.1%
Paris, France cdg1
1,695 3.3%
Stockholm, Sweden arn1
950 1.8%
Portland, USA pdx1
592 1.1%
London, United Kingdom lhr1
532 1.0%
Singapore sin1
304 0.6%
Montréal, Canada yul1
260 0.5%
Mumbai, India bom1
227 0.4%
Sydney, Australia syd1
65 0.1%
São Paulo, Brazil gru1
43 0.1%
Dublin, Ireland dub1
43 0.1%
Tokyo, Japan hnd1
30 0.1%
Seoul, South Korea icn1
24 0.0%
Osaka, Japan kix1
7 0.0%
Cape Town, South Africa cpt1
1 0.0%

What I Took Away From This Experience

After building all eight charts, I arrived at several important conclusions. Here they are — short and to the point.

  • The chart type matters more than the data. The same numbers in different visualizations tell different stories.
  • 3D isn't always better. Volume is great for presentations, but 2D is more honest for precise analysis.
  • Normalization saves the day. When comparing structure rather than absolute values, 100% stacking is unmatched.
  • GSAP + ECharts is a powerful duo. Entrance animations bring charts to life and make them memorable.
  • Data doesn't lie, but it stays silent. Until you present it correctly, it's useless.
  • Traffic geography is strategy. Looking at the Edge Requests map reveals where infrastructure optimization is needed.

Verdict: Should You Learn 3D Visualization in 2026?

Unequivocally — yes. In a world where data has become the new oil, the ability to visualize it isn't a "nice-to-have skill" — it's baseline engineering literacy. Even if you're a backend developer, analyst, or DevOps engineer, charts help you make faster decisions, defend hypotheses more convincingly, and understand systems more precisely.

But there's a nuance. 3D charts are a tool with a sharp edge. They impress, but when used carelessly they distort perception. So learn not just to build them, but also to choose the right visualization type for the task. That is true mastery.

For me personally, this publication became a bridge between the world of data and the world of visual perception. I realized that a chart is a language through which data speaks to humans. And if you learn to speak it, you gain an advantage in any field of IT and AI.