The HTC Thunderbolt Battery Life Issue: A Warning for Smartphones to Come

by Matthew Smith on March 27, 2011

It’s easy to get caught up in the love for ARM. If you read magazines or sites that focus on smartphones, you’re likely lead to believe that ARM processors are the poster-child of power efficiency. Nothing is impossible with this wonderful, easily licensed instruction set. While Intel is soldiering on with its old-fashioned x86 based processors, ARM is pushing forward, making revolutionary products like smartphones and tablets possible.

Wouldn’t it be wonderful if this were true? Unfortunately, as often is the case, reality is bit harsher than the hype – and HTC’s new Thunderbolt smartphone is the perfect example of a problem that smartphones, and the ARM processors within them, are already struggling with.

That problem is the battery. Just ask any HTC Thunderbolt owner about their battery life issues. When used in an area that supports 4G LTE the Thunderbolt’s battery can be drained from 100% to zero by just a couple hours of constant Internet use. Obviously, that’s makes 4G LTE of extremely questionable use. Many people would likely prefer the slower speeds if it means more reasonable battery life, and some users are looking into ways to turn off the 4G LTE altogether (the Thunderbolt doesn’t have that option baked in, for reasons I can’t imagine).

Bashing HTC over this issue is easy, but it also misses the larger scope of the problem. As people look at smartphones more like mobile computers, they’re going to use them in new and different ways. This means not only constant web surfing and frequent downloads but also a demand for more powerful hardware. This summer will be the summer of the dual-core phone, and quad-core ARM processors are already being developed.

This new hardware isn’t without a power cost, however. The rule of thumb with processors is this: if you increase the clock speed of a processor, or increase its size (by adding more cores), you increase its power draw. There are ways to combat the increased power draw, but ultimately it’s a losing battle. For example, a 100 Mhz Pentium Processor had a TDP of 10 watts. Intel’s recent Core i3-530, on the other hand, has a power draw of 73 watts. The Core i3 offers much more performance per watt, as it is literally hundreds of times quicker than the Pentium. But that doesn’t change the fact that the TDP has risen by 63 watts.

An increase in the power draw wouldn’t be a problem for smartphones if battery technology was improving by leaps and bounds. But it isn’t. While processors become faster with every passing year, battery engineers are satisfied if they can eek another 10% out of an existing design.

Without a breakthrough, it seems likely that smartphones are going to run into a battery power wall, beyond which they will have difficulty breaking. A smartphone is not a large device, after all; the 4.3 inch models currently available or likely the largest the market will tolerate. This seriously hinders the size of the battery that will fit into the device, and puts smartphone engineers in a nasty spot.

Again, the Thunderbolt provides a great example. It’s not a small device by any measure; it’s big, thick, and heavy, partially because of the need to house a 1400 mAh battery. Yet that battery simply is insufficient when the HTC Thunderbolt makes use of 4G LTE. A larger battery would help. So would a more efficient processor, or display, or operating system. But HTC, and everyone else, is still dealing with the reality that a phone can only be so large before consumers turn up their noses at it, and that in turn means the battery can only be so large.

ARM, x86, or fairy dust; it really doesn’t matter what you have in your phone. At some point, a more powerful processor will require more power consumption. The Thunderbolt is likely just the first victim of this reality.

Previous post:

Next post: