When HTC announced their latest evolution of their flagship One smartphone, they announced a fairly powerful device, with a Snapdragon 801 and all the performance you could ask for in a high-end smartphone at this level. With a processor like that, and support for 4G pretty much anywhere in the world, there’s no denying that the One (M8) is a speedy smartphone. However, now there are rumors swirling that HTC might be doing a little chest-beating of their own when it comes to how the M8 benchmarks. Every time a new flagship is announced, the benchmarks begin to crawl out of the woodwork and while they’re never really accurate descriptions of what real world performance we’re to expect, they’re definitely a decent yard stick to measure by. How relevant that yard stick becomes when manufacturers start tinkering with things however, is another question entirely.
Here, we’re going to be talking about the AnTuTu Benchmarking app, and the interesting results that some users have discovered while putting the new One through its paces. We’re also going to be talking about the Asian variant of the One M8 which features a slightly faster 2.5 Ghz Snapdragon 801 as that’s the model that Chinese site ePrice used for their tests. When performing a benchmark in the regular AnTuTu app, the One (M8) scores a little over 38,000, an incredible score for any smartphone. However, in the AnTuTu X version of the app, something the developers designed to prevent boosting or targeting benchmark apps for better performance, the One (M8) scored just over 27,000. No matter how you look at things, there’s a big difference there and one that could clearly show HTC targeting benchmarks in order to show off their new smartphone.
It’s worth noting of course that there’s no concrete proof here and benchmarks have never been all that reliable in the past. Still, it’d be a shame to see HTC fooling benchmarking apps in order to deliver a better result, after all the Snapdragon 801 – let alone the Snapdragon 800 – is one of the fastest chips out there, so we question why there needs to be any fakery here. As any of us who have used a benchmarking app before will know, results vary from run-to-run so the discrepancy here might be all that dramatic, but it’s an interesting find nonetheless.