What did real-world 2020 measurements (Tutela) reveal about early 5G's actual performance and global spread?

Early 5G (2020) delivered ~57 Mbit/s average downlink — about 3× the 2G/3G/4G average — far below the 20 Gbit/s theoretical peak. South Korea led deployment by a wide margin, Samsung devices dominated (Apple added 5G late), and NSA dominated over SA.

The reality check (Tutela — crowdsourced data from ~300 million users, 3000+ apps, 200B+ daily data points; dataset Apr–Jun 2020):

• Downlink ~3× higher than 2G-3G-4G; uplink at least ~2× higher
• 5G latency ~18–21 ms (still far from the < 1 ms ideal)

The global picture (2020):

• South Korea was by far the leader in deployment (providers SK, KT, U+); the USA followed (T-Mobile, Sprint)
• Samsung devices dominated; Apple added 5G late
• New York and London were the cities with the highest 5G usage (outside South Korea)
• Switzerland and South Korea led on downlink throughput (AVG ~60 Mbit/s)
• The vast majority of connections were NSA, not SA (e.g., South Korea ~97.9% NSA)

The lesson — theory vs. practice: the 20 Gbit/s headline is a peak under ideal conditions; real early-5G users saw ~57 Mbit/s. A useful, sobering reminder that advertised peak rates and lived experience differ by orders of magnitude — especially in the NSA-dominated early roll-out.

Tip: "~57 Mbit/s real vs. 20 Gbit/s theoretical" is the headline reality check. And early 5G ≈ NSA 5G.