This site may earn chapter commissions from the links on this page. Terms of use.

OnePlus5

For as long equally nosotros've had benchmarks, nosotros've had companies willing to cheat on those benchmarks to make themselves look good. It'due south a never-ending cycle to identify what companies do and how they do it, and nosotros don't often get a clear window into the procedure.

XDA-Developers managed to get their hands on a OnePlus 5, yet, and the results are damning. This is a phone that's designed to deliberately obfuscate its own performance in the name of looking better than the competition. Here'due south how they describe the problem:

The OnePlus 5, on the other hand, is an entirely dissimilar beast — it resorts to the kind of obvious, calculated cheating mechanisms we saw in flagships in the early days of Android, an approach that is conspicuously intended to maximize scores in the most misleading fashion. While there are no governor switches when a user enters a benchmark (at least, nosotros can't seem to see that's the instance), the minimum frequency of the little cluster jumps to the maximum frequency equally seen under performance governors. All little cores are afflicted and kept at 1.9GHz, and it is through this cheat that OnePlus achieves some of the highest GeekBench 4 scores of a Snapdragon 835 to engagement – and likely the highest attainable given its no-compromise configuration with its specific configuration.

The slide beneath shows how the OnePlus v performs if you obfuscate the benchmark information technology's running, and compares it with how the phone behaves in normal circumstances:

Clock throttling when the chip is cheating versus when it isn't. Credit: XDA-Developers

That's the piffling, not-adulterous cluster on top, and the big, cheating cluster beneath information technology. Co-ordinate to XDA, the cheating boosts performance by almost five percent compared with other solutions equipped with a Snapdragon 835.

There are several issues with what the company is doing hither. Offset and foremost, they're driving the phone's SoCs in a style Qualcomm didn't necessarily design for, and that the chip may not be able to handle in the long term. Second, in that location'south no indication these cheating optimizations are applied in any non-benchmark tests. One of the differences between the Turbo Mode frequencies that Intel and AMD offer and this kind of skullduggery is that Intel and AMD are both up front end about the base and max frequency of their processors. Earthworks around in those settings may require some basic understanding of how UEFI works (or how to apply tertiary party tools to accomplish the same thing), but they exercise work and autumn under user control. In that location's no such selection here; users can't plough this feature off in the Settings menu or an equivalent.

And finally, this visitor is advertising its own capabilities as beingness ahead of other Snapdragon 835 devices when it's cheating to get at that place. Over again, we've seen like patterns from motherboard manufacturers in the by, where companies would goose their clock or FSB speeds to run a bit just a tiny bit faster than it otherwise would. But these optimizations don't always work — I've tested high-terminate chips in motherboards before and had them destabilize for no apparent reason until I realized that the UEFI had automatically applied an overclock I wasn't even aware of. Keeping benchmark workloads consequent is critical for testing mobile devices, and bluntly, near companies don't do a very good chore of it.

Given the oestrus and limited dissipation of a smartphone, I've got to agree with XDA on this one. Overclocking a smartphone (which is what's going on here, more or less) is not a smart movement. And the company's decision to effort and pull this play a joke on to expect better in benchmarks isn't going to win it whatsoever points from the states. We didn't like it when Samsung pulled these tricks, and we certainly don't like it at present.

The worst affair about this kind of cheating is that it's done in the proper noun of making a product wait faster at the expense of other factors. Higher clock speeds mean hotter products and less battery life. This doesn't affair much when you piece of work with a desktop, but in mobile devices weight and battery life count for far more than. It's the kind of issue developers should have long since adjusted for. Five percent extra functioning in a few constructed tests of already marginal value merely isn't worth the hassle.