Singtel 5G: Test-driving the red camp's 5G trial network

Testing and Closing Remarks

Testing in the Southern, Central, etc. parts of Singapore

As explained earlier, Singtel's 5G deployment approach is different, so we'd have to pick spots that were sufficiently far apart to be counted as testing in different districts of Singapore while still staying within the 'zone'. To do so, we've travelled to the furthest possible 5G locations within Singtel's domain at the time of writing. Towards the East, we were able to visit Kallang Leisure Park. The westernmost point was Great World City, and the Northernmost was only up to the CBD area (which is technically the Central part of our island). For the Southern tests, we had the chance to try 5G out at Mount Faber and Harbourfront, which were two of Singtel's coverage areas.

We will showcase each location in chronological order - that is, starting from the first set of tests we did, to the last. 

Note: for all locations, we'll use the format of showing you our test spot, followed by the 4G network's speed test scores, and finally Singtel's 5G trial network results.

 

Mount Faber

Would you look at that? The 800Mbps+ real-world results on the Singtel 5G trial network were far better than expected (after all, 1.2Gbps is something we've grown to expect only in near-perfect conditions). We've done several tests around different (walkable) parts of Mount Faber, and there were instances where the speeds broke 900Mbps too. It's unmistakably 5G, even right down to the 5ms latency we've seen. The 4G performance here is also strong, so it's also likely that the area was too good for network strength.

Also, a single press of a test button while on 5G took out 1GB worth of mobile data. Beyond 5G, we also learnt how intangible numbers were capable of inflicting physical pain.

 

Outside VivoCity

We revisited a testing spot where StarHub also staked a 5G claim. Here, the difference in 4G and 5G performance for Singtel were drastic, with the latter giving us nearly 3x the speed across multiple tests. Unlike Mount Faber, the VivoCity area was much more populated, partially explaining the 400 to 460Mbps performance we see - the higher concentration of folks also resulted in more handsets within our range, and that's way more 4G handsets being served by Singtel's base station as we did our testing. 

 

Outside Great World City (Zion Road)

Despite the strong 4G performance here, the trial 5G frequencies still managed to push the speeds even further, resulting in an 11% and higher increase in speed. Perhaps this is our first instance of experiencing the difference between a 5G NSA network relying partially on the 3.5GHz frequency, versus one that uses the 2,100MHz range (in combination with 4G aggregation). 5G network coverage extends at least halfway into Zion Close, where we also got similar speeds down the stretch of road.

 

Outside Leisure Park Kallang

This area's coverage is impressive on two counts - first, the difference is typically 4.5x between 4G and 5G speeds, where 5G offered anywhere from 450 to 650Mbps. Second, it was easy to latch onto the 5G network when did multiple tries across different parts of the mall's vicinity. We got this speed right outside Stadium MRT's Exit B, and we also saw decent 5G speeds in the vast carpark and the open-air area outside of Kallang Theatre. While we did not find the extreme range limits of Singtel 5G's in this area, the total walking distance where we had uninterrupted 5G from carpark to outside Kallang Theatre was at least 250 to 300m.

 

Orchard Road stretch (between Cuppage Road and Bideford Road)

This is the first instance in our testing where we were unable to receive 5G speeds in a location where Singtel has worked on and put up live 5G signals. A major factor that affected our top speeds was the crowd concentration in Orchard close to knock-off time.

How we knew 5G was turned on was also through the test results we got, and why it was important to check it against the 4G network. While we got only 100Mbps~ speeds across the ~500m sidewalk, it was consistently double the 4G network's speed. The low ping in latency also clued us in on the availability of 5G in the area. Also, we understand that there are legislative restrictions around deploying 3.5GHz frequencies in the town area (for the time being), so we're also looking at results produced by the 2,100MHz network. 

 

Telok Ayer Street and Amoy Street

The 5G network around the watering hole area of Telok Ayer saw a consistent 25% increase in speed over its 4G counterpart. While the aggregated 4G network is responsible for 75% of the connectivity speed, the remaining Mbps was further boosted by the possible 2,100MHz frequency in this area. 

 

Our thoughts on Singtel's 5G NSA network and connectivity

On a whole, the speed test results conclusively proved that Singtel's 5G trial network is indeed active in the areas we visited, even if the 5G speeds appeared otherwise in more challenging venues, like Orchard. However, speed isn't the only factor that defined our testing experience, even when Singtel's network was able to deliver close to real-world 1Gbps at its peak.

It was truly a different connectivity experience when we compared our day with the other days spent on StarHub's testing. The ease of latching onto the Singtel 5G network had a strong impact in facilitating the swift collection of multiple test results in one area, while our previous brush with 5G with the green camp was significantly more rigid - 5G were literal spots for StarHub, while 5G for Singtel were across wider stretches or swathes of land.

The counterpoint is the breadth of coverage offered by both telco camps. Singtel's provision is currently limited to (mostly) the Southern and Central areas of Singapore, while we could find 5G spots as far as Woodlands, Bishan, Bukit Batok, and even Pasir Ris via StarHub's trial network. The coverage area is a work-in-progress for both camps, so it's worth revisiting after they've widen the network further, as more 5G NSA base stations come online.

Both rivalling networks were about a week old when we tested them. So, it's quite clear that each telco has a preferred method to deploying 5G over time, and areas of coverage will not be at parity until a much later date. 

We've also noticed the importance of the 3.5GHz spectrum in Singapore's journey towards an accessible 5G landscape for consumers. It partially explains why Singtel was banging so hard on being the first to achieve 1.2Gbps via internal tests, and why they were so excited to be the first to activate the 3.5GHz spectrum out of all our local telcos.

While the specific spectrum we were on are Singtel's closely-guarded secrets, figuring the spectrum out can be partially derived from the test results - in instances where we did not see multi-fold contribution to network speed, chances are, it was in an area where 2,100MHz frequency was activated. Areas like Mount Faber, Leisure Park Kallang etc. were clearly on the 3.5GHz spectrum, because of the massive difference in top real-world speeds. Of course, this will change as more bands are introduced, and it wouldn't be easy to tell apart by inference in the future. Not that the technicalities matter to the average mobile data user, because the defining difference for them is the tangible increase in speed when switching between 4G and 5G - of which both StarHub and Singtel were able to achieve.

As far as Singtel's 2,100MHz spectrum results go, it's comparable to StarHub's 5G trial network speeds, which runs on 2,100MHz and 900MHz. 

As we noted in the StarHub 5G trials, a one-week-old network (no matter which telco) is barely representative of the overall performance. There are still many uncontrollable factors that the real-world speeds we can get when it's finally 'out of beta'. Still, these tests are a promising preview of what's to come for 5G.

If you're not enrolled into Singtel's 5G trial yet, do check out this article to see how you can be a part of the future. 

Additional reading: 

• StarHub 5G: Test driving the brand-new trial network
• StarHub 5G speed tests round #2: How does it hold up against its own 4G?