The test setup
The tester used an HP EliteBook x360 1030 G2 with an Intel Core i7-7600U (a dual-core chip), 16 GB of RAM and Intel HD Graphics 620. That is a seven-generation-old ultrabook with an aging battery. The author gives no battery health or capacity figures.
The three choices live under Settings > System > Power & battery. Microsoft's support page describes them this way:
- Best power efficiency saves battery life.
- Balanced balances performance and battery use.
- Best performance maximizes performance.
You can set each one separately for Plugged in and On battery.
The author kept the underlying Balanced power plan active and changed only the On battery power mode for each run. Brightness, volume and other power settings stayed put. The author also turned off automatic Energy saver, which had been set to kick in at 30% battery. That was a sensible call. Microsoft documents Energy saver as a separate feature that dims the screen and limits background processes when the battery is low, so it could easily have skewed the comparison.
The workload was a four-hour video recording playing while the author did real work. That meant researching in Chrome, writing in a CMS, viewing screenshots in Windows Photos and occasionally editing in Paint.NET. Each run went from about 99% to 10% battery.
Battery results
| On-battery power mode | Runtime (about 99% to 10%) |
|---|---|
| Best power efficiency | 3 hours 2 minutes |
| Balanced | 2 hours 29 minutes |
| Best performance | 2 hours 7 minutes |
- Best power efficiency lasted 33 minutes longer than Balanced, which the author puts at roughly 22% more runtime.
- Best performance ran 22 minutes shorter than Balanced.
- Best performance ran 55 minutes shorter than Best power efficiency. That is about 30% of the efficient mode's runtime.
The author notes that a human-driven workload varies between sessions and is less repeatable than an automated one. Treat these as illustrative numbers, not lab data.
Benchmark results
Cinebench 2026 took nearly two hours on this dual-core chip, and CPU-Z kept stalling. The author fell back on 7-Zip's built-in benchmark. The settings were version 26.02, a 32 MB dictionary, four threads and ten passes, run on battery in each mode.
| Power mode | Compression (GIPS) | Decompression (GIPS) | Total |
|---|---|---|---|
| Best power efficiency | 10.442 | 13.000 | 11.721 |
| Balanced | 10.239 | 12.922 | 11.580 |
| Best performance | 10.132 | 13.016 | 11.574 |
Best power efficiency scored about 1.2% above Balanced. Best performance landed a hair below Balanced, by 0.006 GIPS. The author says these gaps are tiny. With only one clean run per mode, the data doesn't show a meaningful speed advantage for any of the three.
One more caveat matters here. The first Balanced run scored only 9.258 GIPS because performance dropped sharply in the last three passes. A rerun gave the 11.580 figure, with all ten passes consistent. That kind of wobble is a reminder that a single benchmark run on a laptop can be noisy. It could come from thermal behavior or background activity, though the report doesn't say which.
Why this result isn't a universal rule
- Old silicon. A 7th-gen dual-core mobile chip has little headroom to unlock. Newer hybrid-core CPUs may respond differently to power-mode policy. I haven't seen data in this report that settles that either way.
- A narrow benchmark. 7-Zip measures compression throughput. It doesn't show whether Paint.NET filters, Chrome with many tabs or a game feel snappier. The author says so too.
- Possible hidden variables. OEM power-management software, firmware behavior and cooling can all matter. The report doesn't say whether these were controlled.
- The usual caution. Other write-ups note that Best performance can help a strongly CPU-limited game or sustained job on some systems. For web browsing, office work and video, they say the gap between Balanced and Best performance can be small. That comes from PCCurrent's explainer, and it is general guidance, not a measurement from this test.
The takeaway is not that Best performance is useless. It is that "best" describes the policy, not a guaranteed result on your hardware. If you can't measure a gain on your own workload, you are probably just paying in battery life.
How to test your own laptop
- Open Settings > System > Power & battery.
- Under Power mode, choose a setting for On battery. Choose Plugged in separately if you want to compare that too.
- Decide whether Energy saver should be active. Turning it off for the test avoids confusing its effects with the power mode's.
- Keep brightness, volume and the workload as consistent as you can.
- Repeat the run for each mode. Run any benchmark more than once, since one bad pass can skew the average.
If you can't change the mode: Microsoft says a custom power plan may be selected. Open Control Panel > System and Security > Power Options and select Balanced. Community guides also note that the modern selector is tied to the Balanced plan on battery-equipped PCs. That fits the author's note that powercfg /list showed only Balanced.
Don't mix up modes and plans
Microsoft's documentation describes powercfg /list as listing power schemes, also called plans. Windows Settings exposes the power-mode choices separately. The /list output therefore doesn't tell you which modern power modes are available. It only tells you which classic plans exist.
Chasing a drain problem
The original article points readers to a hidden energy report for battery drain. Microsoft documents two related tools, and they do different jobs:
powercfg /energyanalyzes the system for common energy-efficiency and battery-life problems and writes an HTML report. Microsoft says to run it while the PC is idle with no programs or documents open.powercfg /batteryreportgenerates a report of battery usage over the system's lifetime.
Neither is a magic app-blamer, but both are good places to start.
Bottom line
On this EliteBook, Best power efficiency bought 33 extra minutes over Balanced and 55 over Best performance. It did that with no measurable speed penalty in a 7-Zip benchmark. For writing, browsing and video away from a charger, that is a strong argument for the efficient mode. Keep Best performance for tasks where you can see a real gain. Plugged in, the trade-off disappears. Your results will vary with hardware, battery condition and workload, so test your own setup before you copy anyone's default.