Battery Life Part 1: Introduction and Battery Boost Functionality
Battery Life and Power Management
Similar to our GT70 review from a few months back, one of the most intriguing items on the agenda today is how well NVIDIA’s Battery Boost technology manages unplugged efficiency and performance. Certainly the bulk of the most people’s gaming will be done while on AC power, but when that isn’t possible or convenient, Battery Boost is meant to help govern this delicate balance. It accomplishes this by way of frame rate limiting and some internal “voodoo” (as Ryan described it) which NVIDIA declines to elaborate.
Before we go further, we realize the irony of dedicating two entire pages to battery life testing in a review which, as we even proclaimed ourselves, is about a product where battery life really doesn’t matter all that much. But what we’re really interested in here isn’t so much the typical runtime values that we measure most laptops against, but rather the corroboration of Battery Boost’s contributions, which can be handy in a pinch when gaming is desired but outlets just aren’t practical or convenient.
During our GT70 test period, we found the Battery Boost functionality to be helpful, but frequently finicky. We experienced substantial gains in longevity while playing certain games (nearly double with Portal 2; 48% longer in Bioshock Infinite), but absolutely none while playing others (Diablo III, StarCraft 2, Civilization V, World of Tanks). Today, we’ll be testing most of these same products a second time under the same scenarios to see whether the GT72 and its GTX 980M (with newer drivers in tow) manages better results.
Before we get to our testing, let’s first recap the battery situation. Unlike on the GT70, the GT72’s battery is internal—and worse yet, it’s quite difficult to replace. It doesn’t require removal of the board, but it does require removal of nearly every other component accessible from the bottom of the unit, including a second notebook frame panel which is sandwiched between the CPU/GPU heatsink fans and the board. It’s even considerably more complex than the procedure that is followed to replace the GPU itself, since (once again) that’s technically possible thanks to the GT72’s incorporation of a daughterboard that houses the GPU and simply plugs into the board via a special port.
One interesting point to consider regarding performance while gaming unplugged: as Ryan explored in his GTX 980M review, total power consumption is a serious constraining factor when working with high-end mobile gaming hardware. We measured some 200+ watts consumed by the notebook while plugged in, whereas while unplugged, that number would drop to between 80 and 95 watts. That’s a massive discrepancy, and so it begs the question: what is the impact on quality? Ryan’s already answered that for us in his article—but the conclusion bears repeating here.
The updated version of the GeForce Experience software actually divides the stored GFE profile settings for each individual game into two categories: Plugged In and On Battery. While Battery Boost’s purpose is to manage consumption in pursuit of longer runtimes, GFE’s goal is to provide a more consistent level of quality throughout the experience. As such, the recommended quality settings for most demanding games are lower while on battery—but as Ryan concluded, the overall experience is indeed superior to that of the Plugged In settings. That’s because while running at higher settings, the system must throttle its performance to satisfy the substantially reduced TDP constraints of unplugged operation—a last-minute compensation technique which necessarily results in jumpy frame rates and inconsistent performance during gaming. Universally preferable to this unpredictable approach to power management is a “safe” recommended settings profile for each game which proactively reduces the total amount of power likely to be demanded by the game at any given juncture.
Lastly, a quick note on methodology. As with our GTX 880M testing in the GT70 previously, we’ve selected games with sufficient FPS headroom (in excess of 60 fps) to provide Battery Boost some latitude to function as intended. The frame rate limitation settings were left at the defaults (30 fps) for each test, and no other options were modified. The settings used were identical to those in the GT70 review for each game. Finally, we did not interact with the system at all once the test began. This means that Windows’ critical battery warnings kicked the game out of operation near the end of each test—but this shift was consistent across all trials, so its effects on the results as compared to one another are minimal.
Battery Boost Functionality
All right—now that that’s out of the way, let’s get down to some numbers. The first game on our list of trials is Portal 2. Things look awfully familiar to the GT70’s results with Battery Boost off:
A little longer, but an hour of gaming isn’t going to satisfy anyone. Switching it on:
Wow. A result of 2 hours and 10 minutes constitutes a premium of 126% better runtime, or over double the battery life recorded with Battery Boost off. This is somewhat anticipated, however, thanks to the fact that the GTX 980M’s notable performance improvement over the GTX 880M and frugal power consumption under the circumstances. Since it’s already able to do more with less, it’s obviously able to cut even more consumption to reach the easy target of 30 fps.
World of Tanks
First, with Battery Boost OFF:
That’s actually lower than the runtime we received from the GT70, but it’s not completely fair to compare due to the updates and changes to the game’s engine that have taken place. What’s more important is whether Battery Boost provides any discernible benefits; we’ll remind you that on the GT70, we saw absolutely none.
Indeed. A jump to 98 from 61 minutes is an increase of 61%, which is very good. And bear in mind that, yet again, we are not applying the GFE settings in this testing.
The next item in our quartet of games which saw no benefit from Battery Boost during our GT70 testing, Diablo III ought to be a prime candidate for the technology thanks to its low demands and consequent high frame rates on the GTX 980M.
Battery Boost OFF:
Roughly an hour of DIII gameplay will barely get you through a few rifts, let alone satisfy anyone’s addictive tendencies. Switching ON:
Not as dramatic as we had hoped, but that’s still a premium of 63%. Yet again, it’s well worth using if you can handle a limit of 30 fps. Most importantly, it actually works.