Bloatware Kills FPS Performance – Benchmark ft. Norton, 1060 Laptop

By Published August 29, 2016 at 9:49 am
  •  

When we received the new 10-series laptops for review, we immediately noticed sluggishness in the OS just preparing the environment for testing. Even with an SSD, opening Windows Explorer took at least one full second – eternity, by today's standards. It was anything but instant, as a new computer should be, and would prompt outrage from any real-world consumer.

Looking further into the issue, we realized that the system tray accommodated 13 icons of pre-installed software that opened on launch. This included an incessant warranty registration pop-up/reminder, Norton Anti-Virus (the biggest offender on spurious CPU utilization), about three different control panels – because we need multiple paths to one location – and a few other programs.

This, traditionally, is what's known as “bloatware;” it's software pre-installed by the manufacturer that the user didn't necessarily request, and bloats the system's processes to a crawl. Today, we're showing just how profoundly a new system's framerate is dragged down by bloat. Using an MSI GE62VR Apache Pro laptop (~$1600) with a GTX 1060 and an i7-6700HQ CPU (boosts to 3.5GHz), 16GB DDR4, and an M.2 SSD, we're clearly not running Windows on slow hardware. And that's the thing, too – even Windows is slow at the desktop level. Just using the desktop, we'd occasionally spike to ~30% load for no good reason, and frequently hit 100% load during file transfers (thanks, Norton).

For validation purposes, we also ran the same tests on an MSI GE62 Apache Pro with a GTX 970M and i7 CPU. That's one last-gen model and one current model, both clean Windows installs with all the factory-preset software included.

The software pre-installed and launching on boot includes:

  • Norton Anti-Virus

  • Killer Networking

  • SteelSeries keyboard management

  • Touchpad management

  • Warranty registration pop-ups

  • MSI control panel

  • Intel software

  • Microsoft OneDrive

  • Nahimic Audio Enhancer

  • And more

Norton is the biggest offender in terms of active processor utilization; just during file transfers – moving games to the laptop – we saw CPU utilization spike to 100%. The system became unusable. CPU usage often hit 30% sporadically during use, often when Norton decided to scan something.

This is an important topic: Although the readers of our site are largely enthusiast users or informed users, a great majority of buyers won't be. Those people will just assume that the hardware is bad, or that Windows is bad, or that the brand is bad – but in reality, the system can be instantly improved by uninstalling all the bloatware. Even just deactivating launch on Startup improves things. Those tools don't have to go away forever, but can at least be disabled until needed.

Let's look at the impact of this messy OS on gaming performance.

Game Test Methodology

We tested the games using an MSI GE62VR with GTX 1060 and an MSI GE62 Apache Pro with GTX 970M.

NVidia's 372.54 drivers were used for game (FPS) testing on the 1060. Game settings were manually controlled for the DUT. All games were run at presets defined in their respective charts. We disable brand-supported technologies in games, like The Witcher 3's HairWorks and HBAO. All other game settings are defined in respective game benchmarks, which we publish separately from GPU reviews. Our test courses, in the event manual testing is executed, are also uploaded within that content. This allows others to replicate our results by studying our bench courses.

Windows 10-64 build 10586 was used for testing.

Each game was tested for 30 seconds in an identical scenario, then repeated three times for parity.

Average FPS, 1% low, and 0.1% low times are measured. We do not measure maximum or minimum FPS results as we consider these numbers to be pure outliers. Instead, we take an average of the lowest 1% of results (1% low) to show real-world, noticeable dips; we then take an average of the lowest 0.1% of results for severe spikes.

The hypothesis here is that more CPU-bound games will exhibit a bigger performance swing, as the bloatware is taxing the CPU sporadically during uptime.

We tested both systems out-of-box, aside from a GPU driver update, and then ran the “clean system” tests by uninstalling Norton completely, then disabling all unnecessary software in the system tray.

GTA V Benchmark – GE62VR Bloatware & Norton FPS Lag

msi-ge62-bloat-bench-gta-v

Starting us off, the GE62VR with a GTX 1060 pushes these numbers for GTA V at 1080p with our Very High and Ultra settings configuration. Without bloatware, we're running 109 FPS AVG, 78FPS 1% lows, and 70FPS 0.1% lows.

With bloatware, those numbers fall to 90FPS AVG and in the 50s for 0.1% low FPS. Just for the average, that's a change of 20.7% only by disabling all the pre-installed software that comes with the laptop. For higher resolution play, that's a game changer. And the lows are hugely important, too, because we will start seeing visible stuttering with other games – let's take a look at one of those.

Metro: Last Light Benchmark – GE62VR 1060 & GE62 970M Bloatware

msi-ge62-bloat-bench-mll

Here's Metro: Last Light. MLL is one of the more CPU-responsive games we've tested, and is also one of the few games to show sizable performance deltas by toggling hyper-threading/SMT. At 1080p with Very High quality and High tessellation, the GTX 1060 rig runs at 75.7FPS AVG without bloatware, 54.7FPS 1% low, and 51FPS 0.1% lows. With the system running its factory default, including the software, we're seeing a 59FPS AVG and less than half the 0.1% low values. This is a huge hit to performance and will present itself as staggered FPS output. The results also correspond with our discovery of spiked CPU utilization by background processes, since MLL does respond more intensely to such CPU usage swings. There's a major disparity in frametime pacing. More suitably, this also means that these low metrics would be totally incompatible with fluid VR play without dynamic quality adjustments, in the face of MSI's “VR” badge on the system.

msi-970m-bloat-bench-mll

Looking at the GTX 970M unit, we're seeing an output of 40.7FPS AVG with 30FPS 1% lows and 24FPS 0.1% lows with no bloatware. Looking at its performance from the factory, those numbers are almost halved across the board. There's nearly a 2x gain in performance just by removing the clutter, and that's particularly noticeable with the lows. 13.7FPS 0.1% metrics means visible stutter and chop in framerate output, which creates a jarring experience.

Shadow of Mordor – GE62 970M Bloatware Benchmark

msi-970m-bloat-bench-mordor

Here's Shadow of Mordor on the 970M unit at 1080p with Ultra graphics. We're seeing a marginal improvement from 44FPS to 49.3FPS AVG, or approximately 12%.

Overwatch – GE62VR 1060 & 970M Bloatware/Norton Benchmark

msi-ge62-bloat-bench-overwatch

msi-970m-bloat-bench-overwatch

Overwatch is the only game we tested where no major improvement was shown, and that's a result of the game being minimally demanding for this type of hardware. It's also bumping into GPU limits before CPU limits (the taxed resource by the background tasks). The GE62VR 1060 unit is producing a framerate of 147.7FPS AVG with both the clean and bloated systems, but we see marked improvement in the 0.1% and 1% low metrics with the cleaned-out unit. The same is true for the 970M, where performance is most different in the low values. This coincides with our findings that the CPU chokes almost randomly on Norton and other background processes.

Conclusion: Fix It, Vendors

msi-laptop-sidebyside

The take-away here is not that these systems are bad – the hardware's actually pretty performant, looking at the clean unit's results. Unfortunately, though, MSI is ruining its performance with pre-installed clutter, and that's going to hurt their benchmark numbers and, more importantly, user perception of their product. From a competitive standpoint, if MSI – or any vendor – wants to appear better than the next sticker assembly line, they should start with bloatware.

One thing we talk about with system integrators like CyberPower, Origin, and iBUYPOWER is that they all have options to ship the unit with no bloatware at all. This is the correct approach, and should be adopted by MSI. Besides, MSI units are often sold as rebranded SI units. If you were to buy one of these from MSI and then the same unit from an SI who opts out of bloatware, you'd actually see better out-of-box performance.

And MSI isn't the only offender, they're just the one who happened to send us hardware. If we look at the market, Gigabyte ships laptops with XSplit, webcam software, and Smart Dashboard – their version of MSI Dragon Center. ASUS includes Sonic Suite, XSplit, and ROG Gaming Center – another control panel. Some of this stuff is worse than others, of course; XSplit is actually useful and easily disabled, but the control panels aren't really necessary. Norton, of course, is the absolute worst offender to FPS.

Now, of course, there are likely partnerships in place -- financial or otherwise -- where Norton, MSI, XSplit, or other relevant vendors are working together on these implementations. It's not quite as easy as "just stop installing software" for the manufacturers. Unfortunately, their attempt at differentiating a product is working -- but not in the way that they want. The product looks worse as a result.

Editorial, Test Lead: Steve “Lelldorianx” Burke
Video Producer: Andrew “ColossalCake” Coleman
Sr. Test Technician: Mike “Budekai?” Gaglione
Test Technician: Carter Harris

Last modified on August 29, 2016 at 9:49 am
Steve Burke

Steve started GamersNexus back when it was just a cool name, and now it's grown into an expansive website with an overwhelming amount of features. He recalls his first difficult decision with GN's direction: "I didn't know whether or not I wanted 'Gamers' to have a possessive apostrophe -- I mean, grammatically it should, but I didn't like it in the name. It was ugly. I also had people who were typing apostrophes into the address bar - sigh. It made sense to just leave it as 'Gamers.'"

First world problems, Steve. First world problems.

We moderate comments on a ~24~48 hour cycle. There will be some delay after submitting a comment.

Advertisement:

  VigLink badge