The Wi-Fi 7 Router That Overheated: What the Launch Reviews Left Out
Archi4 min read
Contents
- The Test That Wasn't Run
- What the sustained-load run found
- Why the Headlines Missed It
- The bench is not the living room
- Peak numbers are easy, duty cycles are hard
- The failure was gradual, and gradual failures do not go viral
- Lessons for Buyers
- What the vendor got right
- Lessons for Anyone Who Tests Hardware
- The Part the Launch Cycle Hides
The Test That Wasn't Run#
The reviews landed on a Tuesday. Nearly every one led with a number in the tens of gigabits per second, measured on a bench, between two machines, for a few minutes. That number is real. It is also the least interesting thing about a router that has to run in a house for five years.
What the coverage mostly skipped: sustained load. A router is not a dragster. It is a commuter vehicle that idles all day and occasionally carries a full load of traffic for hours. When a small team ran that test on a flagship Wi-Fi 7 unit this spring, the results did not match the launch narrative.
What the sustained-load run found#
The setup was plain. One router, one 6 GHz client, a synthetic transfer held open for three hours in a room at 24 degrees Celsius. For the first 40 minutes, throughput sat near the rated figure. Then it began to slide. By hour two, the 6 GHz radio had dropped to roughly half its opening rate, and the case had climbed past 60 degrees. The firmware was not logging a thermal event. It was quietly reducing power to the radios, which is exactly what thermal management is supposed to do.
The problem is that nobody told the buyer. The spec sheet lists a peak. The launch reviews measured the peak. The behavior after the first hour lives in a support forum thread with 11 replies.
Why the Headlines Missed It#
This is not a story about dishonest marketing. It is a story about measurement habits that have not caught up with the hardware.
The bench is not the living room#
Most launch reviews are produced in a window of days, on a bench, with the router sitting in open air. That environment is generous. A router in a real home sits on a shelf, behind a TV, next to a console that dumps heat into the same corner. The enclosure design that looks fine on a bench becomes the whole story in a cabinet.
Peak numbers are easy, duty cycles are hard#
A single throughput figure is comparable across products, which makes it useful for a buying guide and useless for predicting behavior. Duty cycle is the number that matters: what fraction of time the radios spend at full power before the thermal budget forces a step down. Almost no consumer networking review reports it, because it takes an afternoon instead of 20 minutes.
The failure was gradual, and gradual failures do not go viral#
The router did not die. It slowed. A device that stops working generates a headline. A device that loses 40 percent of its throughput in hour two generates a shrug, because the user blames their ISP, their walls, or the neighbor's microwave. The failure hides inside normal troubleshooting.
Lessons for Buyers#
Three habits would have surfaced this before purchase. , Ask about sustained throughput, not peak. If a review does not mention a long transfer or a heat soak, it has not tested the part you will live with., Look at the thermal design, not the antenna count. Vent placement and heatsink mass matter more than the number of external fins. A sealed plastic shell with no airflow path will throttle, and no firmware update changes physics., Treat the first firmware release as a draft. Vendors tune fan curves and power tables after launch. Check the release notes for thermal or radio power changes in the first 90 days.
What the vendor got right#
To be fair, the throttling was graceful. Connections did not drop. Latency stayed usable. The device protected itself, which is better than the alternative. The miss was in disclosure, not engineering.
Lessons for Anyone Who Tests Hardware#
The deeper failure belongs to the review process. Launch coverage is built around a deadline, and a deadline is the enemy of a soak test. A few practical adjustments:
- Run one long transfer and log throughput every minute. A single CSV file turns a vibe into evidence.
- Measure case temperature with a cheap probe. Ambient plus surface temperature tells you more about headroom than any spec sheet.
- Test in the worst plausible spot. Not a bench. A shelf, a cabinet, the corner where the modem already lives.
- Report the drop, not just the peak. "Peaks at X, settles at Y after 90 minutes" is a sentence a buyer can use.
The Part the Launch Cycle Hides#
Consumer networking has a structural problem. The product cycle rewards a number that can be printed on a box, and peak throughput prints beautifully. Sustained performance does not fit in a bullet point, so it goes unmeasured, and what goes unmeasured goes unimproved.
None of this makes the router a bad product. In a cool room with airflow, it does what the box says. But the buyer who paid flagship money for a device that lives behind a TV deserved to know where the ceiling was, and the launch coverage did not tell them. That gap is the actual failure here, and it will repeat with the next flagship, on the next Tuesday, unless somebody decides that three hours on a shelf is worth more than three minutes on a bench.
More articles