I’ve always been a little suspicious of Windows 11’s Best performance power mode. The name makes it sound like you’re leaving free speed on the table by choosing anything else, and my aging HP EliteBook could certainly use every bit of horsepower it can muster.

But I spend plenty of time working away from a charger, and I wondered how much battery I was burning for that supposedly superior performance.

So I decided to put all three Windows 11 power modes through a proper battery rundown and CPU benchmark to see how much speed I was actually getting for the extra power consumption. I had a pretty good idea of how the results would shake out, but Windows apparently hadn’t read the script.

I put all three power modes through a normal workday

My laptop had plenty to juggle

I tested on my HP EliteBook x360 1030 G2, which has an Intel Core i7-7600U processor, 16GB of RAM, and integrated Intel HD Graphics 620. It’s getting on in years, but it still handles most of my daily writing work, as long as I don’t ask it to perform miracles.

Windows 11 lets you choose between Best power efficiency, Balanced, and Best performance under Settings > System > Power & battery. These are separate from the older Windows power plans found in Control Panel, which confused me initially because powercfg /list only showed Balanced.

To make sure I was comparing the right settings, I kept the underlying Balanced power plan active and changed the modern On battery power mode for each run. I also kept screen brightness, audio volume, and other power-related settings unchanged.

I wanted the test to reflect how I actually use my laptop, so I kept a four-hour video recording playing while researching in Chrome, writing in our CMS, opening screenshots in Windows Photos, and occasionally editing images in Paint.NET. This provided a decent mixture of sustained playback, browser activity, and short bursts of image processing.

I also disabled Windows’ automatic Energy saver, which had previously been configured to activate at 30% battery. That feature can reduce background activity and screen brightness independently, which could muddy the comparison.

I started each run at roughly 99% battery and stopped at 10%, keeping the general workload consistent even though the amount of manual work naturally varied between sessions. This gave me a reasonable picture of everyday battery endurance, with the usual caveat that a human-driven workload isn’t as repeatable as an automated benchmark.

Best performance gave my battery a proper walloping

Nearly an hour disappeared between the extremes

The difference in battery life was way bigger than I expected.

Best power efficiency lasted 33 minutes longer than Balanced, giving me roughly 22% more battery life. That’s a pretty decent bonus when you’re halfway through writing an article, the battery indicator is creeping downward, and the charger is nowhere within arm’s reach.

Best performance, meanwhile, was considerably thirstier. It knocked 22 minutes off Balanced’s runtime and left me with just over two hours before I reached my 10% cutoff. Against Best power efficiency, the difference was a whopping 55 minutes, or about 30% of the more economical mode’s runtime. Nearly an hour of battery life sacrificed at the altar of performance.

Of course, your mileage may vary. My EliteBook has an older processor and an aging battery, so newer laptops might handle things differently. Plus, too many Chrome tabs, background tasks, or heavy photo editing can change results from run to run.

If you are dealing with a mysterious battery drain, you can use Windows’ hidden energy report to find power-sucking apps and bad settings. I kept things simple for this test so I could focus purely on the three main power modes.

Still, losing that much battery made me wonder: what exactly is “Best performance” giving me in return?

The CPU benchmarks threw an unexpected spanner in the works

Three modes, virtually identical performance

I initially tried Cinebench 2026, but my dual-core processor spent nearly two hours wrestling with its rendering workload. CPU-Z wasn’t particularly cooperative either, repeatedly stalling during its benchmark.

Eventually, I turned to 7-Zip, whose built-in benchmark measures CPU performance during compression and decompression. It’s a narrower test than a full system benchmark, but it gave me a repeatable workload without monopolizing my laptop all day.

I ran the benchmarks on battery using 7-Zip 26.02 with a 32MB dictionary, four CPU threads, and ten passes, keeping those settings identical across all three power modes.

Here’s what I recorded:

Higher scores indicate better throughput. GIPS represents billions of instructions per second in 7-Zip’s benchmark rating.

My first Balanced run produced a rather underwhelming 9.258 GIPS, although the score came with a sizable asterisk. Performance had fallen off a cliff during the final three benchmark passes, dragging down the average. I ran the test again, and Balanced returned a much healthier 11.580 GIPS, with all ten passes delivering consistent results.

That’s where the experiment developed a sense of humor. Best power efficiency actually came out on top, edging past Balanced by roughly 1.2%, while Best performance finished a whisker behind it. Apparently, the setting with the grandest name hadn’t received the memo. Those differences were tiny, though, and with only one clean run per mode, there was no persuasive evidence that any of the three offered a meaningful speed advantage.

Windows’ power modes can influence processor behavior and background activity, but how much performance you gain depends on the hardware and the work you’re asking it to do. Best performance may still earn its keep during sustained, demanding workloads that benefit from more aggressive power management. My results simply gave it precious little to boast about.

Compression benchmarks also have limits. They won’t reveal whether Paint.NET responds faster when editing an image or whether Chrome feels more responsive with a dozen tabs open. I didn’t measure those operations individually, so declaring the three modes equally quick across the board would be giving the evidence a rather generous promotion.

Still, the results put a significant dent in my assumption that selecting the most aggressive power mode was an easy way to squeeze extra performance out of an aging Windows PC. On my EliteBook, the battery paid a hefty premium while the benchmark struggled to find anything resembling a performance dividend.

The highest setting can keep its bragging rights

For my everyday writing work, Best power efficiency has earned its place as the default whenever I’m away from a charger. Balanced is there if a particular task calls for it, and Best performance can come off the bench when there’s a good reason to give it a run.

Until then, I’m happy to leave the most aggressive setting alone. My aging EliteBook already has plenty of miles on the clock, and I’m in no hurry to squander nearly an hour of battery life chasing a performance boost that couldn’t even make a convincing appearance in the benchmarks.