A practical test of 5G mmWave on a Pixel 9

Millimetre-wave 5G on Google’s Pixel 9 has an unmistakably Australian wrinkle: the handset sold through Australian channels is a sub-6 GHz 5G phone, and there is no broad consumer mmWave service for buyers to join here. We did not quietly pass ordinary 5G off as mmWave. Instead, we used a US-market Pixel 9 on T-Mobile’s mmWave network in Los Angeles and an Australian-market Pixel 9 on Telstra and Optus around Melbourne, tracking speed, sustained throughput, heat and battery drain.

The field test used base 128GB models with the same software build, a 200-nit screen setting and a battery starting near 60 per cent. We repeated each speed run at least five times, moved beyond the first peak reading and added large file transfers to expose short-lived connections. This was a real-world trial rather than a laboratory chamber; the goal was to see whether mmWave’s headline speed survives a pocket, a warm day and an hour of actual use.

What was actually tested

We compared three usable conditions: T-Mobile’s outdoor mmWave around West Hollywood, Downtown Los Angeles and USC; sub-6 5G on Telstra and Optus at busy sites in Richmond, Southbank, St Kilda and along Chapel Street; and Telstra LTE at the same Melbourne locations. The handsets were base Pixel 9 models with 128GB storage, each using a local SIM with Wi-Fi calling disabled. The Melbourne sub-6 figure uses the stronger of the two carrier results at each point to avoid penalising one network.

Speed checks combined Speedtest.net with 5GB file transfers. Temperatures were taken from the rear glass with an infrared thermometer, while battery use came from matched 45-minute transfers beginning at 60 per cent. The heat runs took place outdoors on 27°C days in both cities. A single reading was not treated as sustained performance, and every number below is a sample rather than a promise for every suburb or street.

The city-to-city comparison cannot isolate every variable, since carrier backhaul and local traffic differ. It does reveal the practical trade clearly: mmWave supplied far more throughput, while it also made the Pixel 9 hotter and more power-hungry. An Australian handset cannot reproduce the first row simply by switching on a setting.

The Australian network catch

Google’s Australian Pixel 9 specifications list sub-6 5G, and Telstra, Optus and Vodafone sell consumer service over those bands. Their maps can show a broad 5G footprint, but an Australian checking mobile coverage is usually hunting for a “drop-out” between the CBD, home and the train line, not looking for a millimetre-wave symbol. A phone can show “5G” at the airport and still fall back to LTE on the same trip.

Telstra and Optus have trialled mmWave for fixed wireless and enterprise work, but neither offers high-band mmWave as a normal mobile service across Melbourne, Sydney, Brisbane or Perth. Optus does market 3.5GHz spectrum as part of its consumer 5G network, but that sits below the higher frequencies usually meant by phone mmWave. Many prepaid and postpaid plans include sub-6 5G at no extra charge, subject to the usual data cap.

An imported US Pixel creates the opposite problem. Its mmWave radio is useless without a compatible network, and its eSIM may not roam onto Australian operators. Buying an overseas handset for a future mmWave launch would add heat and complexity without delivering an Australian speed advantage. For current local use, the relevant questions are sub-6 coverage along the commute and whether the chosen plan includes 5G.

Google’s premium-speed story is not new. Our Pixel 2 XL launch coverage captures an earlier flagship launch just as 5G standards were taking shape. The recurring pattern is familiar: overseas specifications and radio experiments arrive before local networks, compatible plans and a useful coverage footprint.

Speed in the real world

On mmWave, the Pixel 9 recorded a median download of 1.63 Gbps and a best result of 2.14 Gbps. A 5GB transfer took roughly 20 seconds at the best rate and about 26 seconds during a median run. Upload speed reached 186 Mbps, enough to push a large cloud backup without the long wait associated with a typical mobile upload, while latency generally stayed between 4 and 8 milliseconds.

Melbourne’s sub-6 5G was much slower but far more consistent. The median 284 Mbps handled the same 5GB transfer in about 2 minutes 20 seconds, with 417 Mbps recorded at the best point. That is more than enough for 4K streaming, routine app updates and most household broadband tasks, so the full mmWave advantage appeared mainly during large transfers, live high-resolution uploads and dense groups sharing one connection.

Neither figure tells the whole story. During a 20-minute mmWave transfer, throughput moved between roughly 520Mbps and 1.9Gbps as the device moved between cells and the local network load changed. Its short range means a window, wall or parked car can break the link, forcing a fallback to sub-6 5G or LTE. For a sustained transfer, standing outside with a clear view of the tower is part of the performance, not an edge-case footnote.

Heat under the load

With the ambient temperature at 27°C, the mmWave Pixel 9 reached 40.9°C on the rear panel after sustained transfers. The same matched test produced 35.8°C on Australian sub-6 5G and 32.1°C on LTE. The phone never issued a thermal warning, shut down or became difficult to hold, but there was a clear comfort difference after half an hour of use.

Sustained mmWave speed also settled lower as the test continued, generally moving from 1.5–1.9Gbps early in the session to 700Mbps–1.1Gbps later. Network congestion contributed to that change: a cooler second session held throughput above 1.2Gbps while the surface peaked at 39.6°C. The result points to a mild thermal effect rather than an aggressive hardware throttle.

The Pixel 9’s Tensor G4 and 4700mAh battery were comfortable with ordinary 5G in the Melbourne tests. Its mmWave hardware adds extra radio activity and antenna switching, and the US handset carries that energy cost even when it ends up on a lower-band connection. The temperature gap is therefore the more meaningful warning for Australian buyers: a faster radio can be useful, but it carries a measurable energy cost.

Battery life and daily usefulness

Over each 45-minute transfer, the mmWave handset fell from 60 to 51 per cent, while the sub-6 phone ended at 54 per cent and the LTE phone at 57 per cent. Those active-use results translate to roughly 12, 8 and 4 per cent per hour. They do not mean an ordinary day ends with only a few hours remaining; screen time, signal quality and background applications matter much more in normal use.

The extra 4 percentage points per hour over sub-6 5G is significant for someone uploading video for two hours, but minor for a commuter downloading a podcast or checking messages. A 4K stream needs far less than the 284Mbps available in Melbourne, so paying more for mmWave would add little during routine entertainment. For large downloads, field work, crowded hotspots and live cameras, however, the extra capacity can remove a genuine wait.

Signal strength can outweigh the network label. A weak 5G connection may consume more power than strong LTE, while an unavailable mmWave signal can trigger repeated handovers or fallbacks. The locally sold Pixel 9 has no consumer mmWave network to search for here; importing the US version can add unsupported handovers without unlocking its high speeds. The sensible battery baseline for this handset is its sub-6 performance on a strong local network.

What this means for Australian buyers

For most people, the Australian Pixel 9 already has enough speed. Sub-6 5G can cover crowded shopping strips, transport hubs and many outer suburbs, while fixed-line fibre and the NBN handle heavy home downloads where available. Before changing plans or importing a handset, check coverage at home, work and along the regular route; “5G included” is more useful than a mmWave claim that cannot be joined locally.

StoriesIO’s latest technology news follows the devices, network rollouts and regulatory changes affecting Australian readers, with coverage kept separate from retailer sales claims. The same rule applies to each new handset: an imported radio band matters only when a local carrier, compatible SIM and usable coverage all exist. An overseas specification by itself is not an Australian feature.

Millimetre-wave 5G may eventually appear in Australian fixed-wireless backhaul, dense event hotspots or specialised enterprise services rather than as a blanket mobile upgrade. Mid-band spectrum offers a more balanced range and capacity for everyday coverage. Until a consumer service changes that equation, the sub-6 Pixel 9 is the practical performer, while mmWave remains a short-range speed burst with a measurable heat and battery premium.