MSI GT70 2PE Dominator Pro Gaming Notebook Review: Appropriately named
NVIDIA Battery Boost Functionality, Battery Life
Now here’s where things get interesting. Clearly, battery life while gaming isn’t of paramount importance in most scenarios to most users, but seeing as that’s the area in which the GT70 really excels (apart from the obvious advantages in terms of raw performance), we have a perfect opportunity to further test NVIDIA’s Battery Boost technology to see just how much (if at all) it manages to improve battery runtimes in particular games.
A quick note about methodology: our Battery Boost testing spans five different games where frame rates are expected to exceed 30 frames per second on the recommended settings for the 880M (and at the native resolution of the panel, 1080p). This way, we’ll allow the 880M some latitude by which to conserve energy through the reduction of the frame rates (which is its principal approach). For each test, we limited the frame rate to 30 frames per second while Battery Boost is enabled—the minimum rate which we would judge to be comfortably playable for most gamers. Brightness was set to a consistent value as close to 150 cd/m² as possible.
One last note: we also did not interact with the computer during the battery testing so as to keep the benchmark conditions as consistent as possible. A side effect of this is that once Windows began warning us of low battery life, the consumption drops for the remainder of the test. This transition is clearly visible on the graphs, where the battery charge slope is greatly reduced.
First up, let’s take a look at Portal 2 with default recommended settings (1080p, all settings High). We of course switched off the option within Portal 2 to automatically limit frame rates on battery power, as we’ll be letting the nVidia software handle that instead.
First, with Battery Boost OFF (no frame limiting):
That’s just 51 minutes of total runtime, which is about what we’d expect to see given the considerable power requirements of the GT70. But let’s switch on Battery Boost (30 FPS limit) and see what transpires:
Yep, you read that right: 1 hour, 38 minutes of total runtime, which is an increase of 47 minutes, or in other words, nearly double (92% longer). This is obviously on the high end of what Battery Boost can achieve, but it’s still a great result and an indicator of just what’s possible.
To help visualize the two results:
World of Tanks
Next, we’ll try World of Tanks, which has quickly become one of the most popular free-to-play games in existence. It’s well known (at least at the time of this writing) for its relative inefficiency in terms of CPU usage, seeing as it’s only able to take advantage of a single core.
Battery Boost OFF:
Battery Boost ON:
That’s a measly 3 minutes difference, which is essentially negligible. So it appears in World of Tanks that our Battery Boost settings didn’t affect anything.
Bioshock Infinite runs at a pretty nice clip on the GT70, and that’s at max settings, too. We’ll get into more performance specifics on the next page; for now, let’s stick with the plan.
Battery Boost OFF:
63 minutes, which is right around what we saw from World of Tanks.
Battery Boost ON:
93 minutes total, which translates to a runtime premium of 48% longer. Not quite as dramatic as our Portal 2 results, but pretty good nonetheless.
Diablo III, StarCraft 2, and Civilization V
Interestingly, in much the same vein as our World of Tanks findings, our Battery Boost tests on each of these three games showed either no change in battery life or negative results. It’s unclear why this was the case; in each situation, the setup was identical: we ran Battery Eater Pro, set it to Idle and to begin logging when unplugged, dialed back the brightness to the same preset, and unplugged the system. Then we launched the game, navigated to the spot where the benchmark was to begin, and let the system idle at that location until it died.
It appears that Battery Boost only really works in specific circumstances. The amount of frames per second “cut” from the rendering equation doesn’t appear to be directly correlated in all circumstances, either. A great example of this would be Diablo III, where the GT70 can easily pump out 81 frames per second at 1080p resolution on all settings at their highest (as we’ll see later in this review). In spite of that, setting an FPS limit of 30 in the Battery Boost menu does nothing to improve the runtime.
This means that, at least currently, many games do not appear to be supported and will not experience any additional longevity as a result of its use. On the other hand, NVIDIA has said that some software-level adjustments take place on the fly during a game’s operation—so it’s possible that in some games it may just take some time to adjust. Furthermore, as additional updates to the GeForce Experience software and the GeForce drivers are released, we should see improved support. In our case, however, we found no benefit in any of these three titles specifically.
Traditional Battery Tests
Next up, let’s have a look at our traditional battery tests. First, the Battery Eater Pro Classic Test:
This test is performed at the same level of brightness as our other gaming battery benchmarks. With just 1 hour and 13 minutes of runtime, it’s hardly a great result; but it’s also consistent with what the rest of the tests have shown.
At the opposite extreme, the Reader’s Test (same settings otherwise):
4 hours and 49 minutes during the Reader’s Test is considerably more worrying than our Classic Test results, and is certainly a premonition of what we’ll see during the web surfing test.
Speaking of which, we’ll turn to that next. Here, as always, we refresh a static web page at regular intervals under the same brightness and power settings as before to see how long the notebook can last:
3 hours and 14 minutes—a poor result by any metric, though again, this is a desktop replacement first and a notebook second. Granted, it’s not much worse than what we measured on the GT60, but the point is, it’s still worse, which is unfortunate. No doubt partially to blame is the larger 17.3” LCD screen as compared to the GT60’s 15.6” panel.
As always, we exclude the web surfing test from our final comparison due to possible differences in browser rendering, page content, etc. Here’s a look at how the Reader’s and Classic tests stack up:
All things considered, just a quick note about the hefty weight and lackluster battery life of the GT70: truly, does it matter all that much? Sure, it’s certainly inferior to nearly all other modern notebooks, even including most gaming PCs. But ultimately, the purpose of the GT70 isn’t going to be gaming on your lap in the car or in a plane most likely. Rather, it’s a desktop replacement—and in this regard, its greatest appeal will be that of portability around the home, to allow for gaming in any room and without the need for a bulky multipartite setup. Stated another way: if you intend to use the GT70 for any purpose other than home-based gaming, you’re probably looking at the wrong machine. This isn’t designed to be a notebook that will easily occupy a bag on a long flight or to and from classes. That’s what an Ultrabook is for, or if you’re really in the market for highly portable gaming, maybe you want a GE40; the GT70, in contrast, is a beast.