What did the ThinkPad P1 Gen 9 actually measure?
Notebookcheck published its review of the Lenovo ThinkPad P1 Gen 9 on 5 November 2026. Three days later, it reported the result that made the laptop notable: 22 hours and 38 minutes of Wi-Fi browsing. The test used Microsoft Edge at 150 cd/m² screen brightness with Wi-Fi connected.
Under sustained load, the same machine lasted 2 hours and 2 minutes. That contrast matters. A workstation can look all-day capable in an office workload and still drain quickly when the processor and graphics card work together.
The review unit was a 16-inch model with Intel's Core Ultra 7 366H processor. That chip has 16 cores in total: four performance cores, eight efficiency cores and four low-power efficiency cores. Notebookcheck lists a sustained power limit of 65W and a short-burst limit of 80W.
Lenovo paired it with an Nvidia RTX PRO 2000 Blackwell laptop GPU, 64GB of LPCAMM2 memory and a 1TB PCIe 5.0 SSD. The GPU is capped at 75W. The battery is a 90Wh pack, and the charger is a 140W USB-C unit.
The review unit weighed 1.947kg. Lenovo's specification sheet says the line starts at 1.80kg, so weight varies with the screen and other parts. The machine is thin for a workstation at 21mm, but it is not an ultraportable.
Notebookcheck called the result the best battery life it has measured for any ThinkPad P1. It said the figure was more than double the line's previous record. The review also gave the laptop an overall rating of 91%.
One caveat belongs at the top. This is a single review unit, tested by one lab, using one scripted workload. That does not make the number wrong. It does mean the number describes a configuration, not every P1 Gen 9 a buyer can order.
Lenovo's own specification sheet shows why configuration matters. The P1 Gen 9 can be ordered with three very different 16-inch panels, up to 96GB of memory and up to 8TB of storage. Each choice changes power draw.
Why is 22 hours unusual for a Windows workstation?
The ThinkPad P1 has always combined a powerful Intel H-class processor with a discrete Nvidia GPU. That combination bought performance for CAD, rendering and data work. It also punished battery life, even when Lenovo fitted a large battery.
Notebookcheck's own history shows the pattern. The 2019 P1 lasted 8.9 hours in the same Wi-Fi test. The Gen 7 reached 9.9 hours, and the Gen 8 managed 5.5 hours.
The Gen 8 figure is the important one. It had the same 90Wh battery as the Gen 9, so battery capacity alone cannot explain the fourfold difference. Something else changed in the platform, the panel or both.
The wider workstation class is not close to 22 hours. Notebookcheck's current average is 10.1 hours across 21 recent machines, with individual results ranging from 3.58 to 22.6 hours.
Other 16-inch laptops in the same comparison table land between 10 and 19 hours. A Dell XPS 16 with a Panther Lake Core Ultra X7 358H and an OLED screen reached 10.3 hours. A Samsung Galaxy Book6 Ultra with a Panther Lake chip and an RTX 5070 reached 16.8 hours.
Apple's MacBook Pro 16 with an M5 Pro reached 18.5 hours in the same test. The HP ZBook X G1i 16, which uses an older Arrow Lake processor and an IPS panel, reached 12.5 hours.
Those figures put the ThinkPad at the top of the table rather than in the middle. That is why a workstation review became a battery story. It is also why the result needs context before anyone treats it as a promise.
Workstations have historically been the worst place to look for endurance. Buyers wanted cores, memory and a certified GPU, and they accepted a short run time as part of the deal. A 22-hour result challenges that assumption, at least for light work.
How much of the gain comes from Panther Lake?
Panther Lake is Intel's Core Ultra Series 3 mobile family. It is the first compute platform built on Intel's 18A process, and Intel launched it at CES 2026 in January. The company said the family would power more than 200 laptop designs.
Intel's most quoted battery claim was 27.1 hours of Netflix streaming. That figure came from a Core Ultra X9 388H inside a Lenovo IdeaPad reference design. A reference design is a platform built to show what the chip can do, not a laptop on a store shelf.
The company also claimed up to 60% better multithreaded performance than Lunar Lake at 25W. It claimed more than 77% faster gaming across 45 titles. Both claims compare specific chips in specific test setups.
The pitch was Lunar Lake efficiency at a higher performance level. Lunar Lake had already moved Intel's thin-and-light laptops toward much longer battery life. Panther Lake had to keep that efficiency while adding cores and a larger GPU.
The ThinkPad P1 Gen 9 is a useful test of that pitch. It uses a 65W H-class chip and a discrete GPU, which is a much heavier workload than a thin-and-light design. If efficiency held here, it would hold in easier machines too.
There is supporting evidence beyond the battery figure. In PCMag's testing, a Core Ultra X7 358H was about 30% faster than a Lunar Lake Core Ultra 268V on PCMark 10's Modern Office test. That suggests the newer platform is not trading all its speed for endurance.
Notebookcheck also measured idle power of about 5.9W with an external monitor attached. Low idle draw is central to a long browsing run, because a scripted web test spends most of its time close to idle.
Panther Lake therefore explains part of the gain. It does not explain all of it. The rest sits in the display, the battery size and the way the test is run.
How much comes from the screen?
The review unit had a 16-inch 1920x1200 IPS panel with a matte finish, 500 nits of brightness and a 60Hz refresh rate. Lenovo's specification sheet labels this the low-power option. It is the least demanding screen in the line.
Lenovo also offers a 3840x2400 IPS panel at 800 nits. The third option is a 3200x2000 tandem OLED touchscreen with a 40-120Hz variable refresh rate and a peak HDR brightness of 1,500 nits. Both alternatives draw more power than the 1200p panel.
The comparison with the Gen 8 is not clean. That review unit had the 3200x2000 tandem OLED, according to Notebookcheck's table. The 5.5-hour result and the 22.6-hour result therefore differ in chip, panel technology and resolution.
Notebookcheck acknowledged the point in its review. It wrote that the basic WUXGA panel of its review unit "surely helped to increase the battery life." It also said it remained to be seen whether the result holds with higher-resolution screens.
PCMag ran a cleaner experiment with two Dell XPS 14 laptops in the same chassis. One had a Core Ultra X7 358H and a 2880x1800 tandem OLED touchscreen. The other had a Core Ultra 5 325 and a 1920x1200 display.
On PCMark 10's Modern Office test at 100 nits, the OLED model lasted about 14.5 hours. The 1200p model lasted more than 33 hours. PCMag said the displays contributed to the discrepancy.
Those machines also differ in processor, so the test is not a pure display comparison. It still shows how far battery life can move when resolution and panel technology change. The gap is measured in tens of hours, not minutes.
OLED panels add another variable. Their power draw depends on what is on screen, because dark pixels use less energy than bright ones. A browsing test with mostly white pages is not the same as a dark-mode video test.
Resolution matters too. A 3840x2400 panel has four times as many pixels as a 1920x1200 panel. Driving those pixels, and lighting them to a usable brightness, costs energy even when the processor is idle.
The practical lesson is simple. A 22-hour headline from a 1920x1200 low-power panel does not transfer automatically to the OLED configuration. Buyers who want the better screen should look for a test of that exact combination.
What the 22-hour headline leaves out
The first omission is load behaviour. The same laptop lasted 2 hours and 2 minutes under sustained load, against 22 hours and 38 minutes of browsing. The ratio is roughly one to eleven.
That gap is normal for a workstation with a discrete GPU. A 90Wh battery drains quickly when the processor and graphics card run together. The 140W charger exists for the same reason.
The second omission is heat. Notebookcheck listed high chassis temperatures under load as a con. It also noted that the GPU is limited to 75W, which is modest for a workstation-class graphics chip.
The third is test brightness. The Wi-Fi test ran at 150 cd/m², which is dimmer than many people use indoors. Raising brightness shortens runtime, and outdoor use will shorten it further.
The fourth is price. Notebookcheck reports a starting price of about $3,200 in the US, with the tested configuration around $5,100 and a list price of €5,300. That is premium workstation money.
The fifth is the workload. A scripted browsing test is light on the GPU. Video calls, large spreadsheets, code builds, image exports and renders will all pull the result down.
The sixth is battery age. Lithium-ion packs lose capacity over time, and a 90Wh battery that starts at 22 hours will not stay there for the life of the machine. Lenovo sells the pack as customer replaceable, which helps.
None of this makes the result false. It makes the number conditional. A conditional number is still useful, as long as the condition is printed next to it.
Who benefits, and what should buyers check?
The clearest winners are people who carry a workstation but mostly write, read, email and join calls. Field engineers, consultants, sales teams and writers fit that profile. For them, a full day away from a socket is a real change.
Light CAD and data work also benefit, because the machine can stay on battery through a working day. The 140W charger and its 419g weight remain part of the travel kit, but they can stay in the bag more often.
Buyers who need colour accuracy or a high-resolution OLED should expect less. The same applies to anyone rendering video or training models on the laptop itself. Those workloads will behave like the two-hour load test, not the browsing test.
A sensible checklist starts with watt-hours, then panel type, resolution and refresh rate. Next come the tested brightness and the exact workload behind any claim. Those four details explain most of the variation between reviews.
It also helps to check the graphics power limit, the charger size and the measured weight. Those details decide whether a laptop is portable in practice, not just on a specification sheet.
Compare configurations rather than brand headlines. The same model can differ by many hours depending on the screen and chip inside. Lenovo's three panels are a clear example of that.
There is also a value question. A $3,200 workstation with a 1200p panel is not the right buy for someone who mainly wants a bright OLED screen. It is the right buy for someone who wants a quiet, durable machine that lasts through a day of light work.
- •Check the watt-hours, not just the claimed hours.
- •Identify the panel: resolution, IPS or OLED, and refresh rate.
- •Match the review configuration to the one you plan to order.
- •Read the tested brightness and workload before trusting a headline.
- •Check the graphics power limit and the charger weight.
What happens next?
More Panther Lake workstations are now shipping, including models from HP, Dell, Samsung and Lenovo. Their battery results will show whether the ThinkPad is an outlier or the start of a pattern.
AMD's Ryzen AI 400 mobile chips and Qualcomm's Snapdragon X2 Elite are the immediate competition. Both will be judged on the same mix of performance, battery life and app compatibility.
Intel's 18A process is also under scrutiny, because it carries the company's manufacturing ambitions. A strong efficiency result in shipping laptops helps that story more than a reference design does.
For buyers, the near-term question is narrower. Can a Windows workstation last a full day in the configuration you actually want, with the screen you actually need?
The ThinkPad P1 Gen 9 suggests the answer is yes for the low-power panel. For the OLED version, the answer is still unproven. The next round of reviews should settle it.