Recent headlines have talked much about the future of smartphones. We hear the phrase “flagship” banded around, a moniker typically reserved for a company’s “headline” product release, used to lead the company flag forward into battle in the marketplace. Back in 2009, when Android first began to gain a sensible level of traction, a flagship was a manufacturer’s main handset for the year.
First, a brief history lesson
OEMs had clear product ranges, and there was a clear “leader”. That was the flagship – the big brother of that year’s phone releases. Each manufacturer would compete to be that year’s best flagship – and we even recently did a nice run down on the flagship history of Samsung, Sony, and HTC.
Look at devices over the years, though, and you’ll see this for yourself. The HTC Hero was probably the first “mainstream” Android device that you could buy easily at launch (previous devices like the G1 were not so easy to get hold of, unless you went out of your way to try to buy one). The Hero shipped in June 2009 with a 528 MHz single-core CPU, and only 288 MB of RAM (a non-negligible portion of which was reserved for graphics, if memory serves your aging editor correct). Only 9 months later, in March 2010, HTC announced the Desire, with its 1 GHz CPU, and 576 MB of RAM!
Taking a step back here, we saw what amounted to almost double the CPU, and double the RAM. In a 9 month period, across a single product generation. These kinds of performance improvements offered customers genuine progress, and made an upgrade compelling. With hindsight, 9 months is a rather short interval between flagships – users are unlikely to be willing to part with their cash twice in 12 months for a new handset, but nonetheless, this model continued, as 9 months was the shortest cycle they could manage, while having enough time to polish a product off, and move onto the next one, before marketing took over the product to sell it to customers.
Then things went downhill
The one thing flagship devices had in common was their price. These were not low-priced handsets. Manufacturers and carriers were almost certain at the time that they were pricing a large number of customers out of the market. They were aware that smartphones should have been selling more units, but customers were being expected to either pay a large amount more for handsets than they were used to (if buying a handset on its own), or to take out a monthly contract (possibly with an upfront contribution to the handset), with a higher monthly fee than they were used to. Word filtered through to OEMs, and their product teams were told that the customers (carriers) wanted to have handsets they could retail at around half the price.
When buying my own first smartphone, I recall facing a decision between the expensive HTC Hero, and a much cheaper (around 1/3 of the price) carrier-rebadged version of an anonymous Huawei handset. Ultimately, I ended up with the Hero, partly thanks to my ability to drive a hard bargain with the carrier, and partly because of a small subforum on a website that I had found. That website was XDA-Developers, and it led me to realise that there was more to a phone than specifications and price.
In the months that followed, though, the main manufacturers started to realise they were at risk from the cheaper handsets, and their product development resources ended up split, with some working on the traditional top-of-the-range devices, and others working on lower-priced models, for the more cost-sensitive sectors of the market. At some point in this process, though, someone had the idea (which I feel was terrible, but which I can only presume they thought was brilliant) to create more of these lower-priced handsets, with different appearances and innards, to cater for all of the possible different price-points on the market. This led to the usual flagship devices continuing as before, but with an increasing number of rushed, poor quality (and frankly junk) devices brought to the market in a rush, by OEMs eager to reach price-conscious buyers.
This led to the market being watered down. There was still the “big” product being launched each year, with impressive specifications to match its impressive price tag, but there were growing numbers of lower-end devices being released. Often, to capitalise on the “fame” of the flagship, OEMs sought to associate their new, lower-end devices, with the existing flagships by using similar names. This is what has led us to seeing devices like the Galaxy S3 Mini which, despite its physical appearance, has a 1 GHz dual-core NovaThor CPU with almost nothing in common with the quad-core 1.4 GHz Exynos CPU of the “real” Galaxy S3.
Now back to today!
Today, flagship devices continue to exist. HTC’s latest device in the One range would be their flagship. Sony’s Z3+ (or Z4, depending on market) would be their flagship. Samsung’s would be their Galaxy S6. And LG’s would be the G4. It’s rather intuitive and obvious to those in the industry, but there’s no official designation from each company, showing what their flagship is. That’s because the market is now saturated with phones, as we just discussed. The market is now awash with so many different handsets that companies are struggling to build a viable marketing campaign for all their different product ranges. Even after HTC said they would stop releasing endless mid-range devices, they continued to do so, apparently unabated. OEMs flooded the market with cheap, commodity handsets.
This was naturally good for the market, as consumers wanted cheaper handsets, and the free market simply ensured they had sufficient supply (and competition) to drive down prices. Such highly competitive marketing, along with high hardware costs (and low margins) led to problems which we continue to see today, however. Software Updates…
Yes, software updates. To the average user, these are relatively unimportant, until something stops working on their phone, and they either blame their last update, or decide their device needs an update to fix it. Updates are expensive, and require significant engineering resources to get right. Getting software right is hard, as OEMs are expected to make software which works on every one of the hundred-plus countries in which their device was sold. The problem is that the engineers are already working on the next mid-range handset, about to be churned out. This makes it very difficult for OEMs to produce software updates for their older phones. This rears its head when major security issues are found, and manufacturers take months to send out a tiny update to fix the bug (and many will never even release such an update, simply deciding it’s too much effort!)
This has led to hardware becoming a commodity, however – companies now ship so many devices, and at such a high pace, that there’s little special about a device launch – there is another one coming up shortly! The sheer competition has turned the very handsets we use into commodities, rushed out of the hands of the engineers and designers, so the next one can be launched. And this leads to devices which are near-identical being released, one after the other.
An Example – Sony
I could have chosen almost any OEM here, but I chose Sony, since the stagnation of their product line is probably clearest to show. Let’s look at their last three flagships. The Xperia Z2 was released in April 2014, with a Snapdragon 801 (MSM8974AB CPU), which is quad core, 2.3 GHz. The Xperia Z3 was released 5 months later, in September 2014, and boasted a Snapdragon 801 (MSM8974AC), with 4 cores at 2.5 GHz. If you blink and re-read the last sentence, you’re not alone – to spare you the effort of seeking a magnifying glance, the different between these two handsets is a slightly different stepping of CPU, and a boost of 200 MHz (on a device which is already at 2.3 GHz). The devices are otherwise pretty much identical, to the extent that I’ve seen a Z3 go into a Z2 case and look fairly comfortable.
Sony Xperia Z4 and Xperia Z3
With the Z3+, Sony moved to the Snapdragon 810 radiator, ahem, CPU. To be precise, the MSM8994. This is a bit different, having 4 low-power cores, and 4 regular 2 GHz cores. Otherwise, the device is pretty much the same. There’s really very little difference between any of them. I was tempted to place a link here to some comparison between the devices based on synthetic benchmarks. Instead, I chose not to, and to simply point out that there’s very little difference between any of the devices. And arguably the Z3+ (with its inbuilt egg-fryer capabilities) would be slower under certain conditions (such as those that don’t involve sitting idle!).
Putting aside those yolks (sorry, last one, I promise), there is a serious point here – there is almost no compelling difference between 3 subsequent generations of Sony devices. Why would someone with a Z2 upgrade? What does the new device have, which the old device doesn’t have? With hardware having reached commodity status (as discussed above), we’re at the point where the answer frankly is “nothing”.
Nothing Different
2015 has been a somewhat boring year for phones. The biggest excitement to the market probably came from Samsung, who decided to grind the corners off the glass of an S6, and call it the S6 Edge. While there’s no doubt customers find it a nice-looking phone, I’ve heard many argue it’s a gimmick. It’s certainly not something which interests me. Yet in a stagnant market, it’s captured the interests of customers, and led to record demand. Indeed, I believe at one point, Samsung’s glass-edge-removal process couldn’t keep up with the demand!
Why? Because the software is rushed, and everyone is focusing on the commodity hardware. Nobody is taking the time to do some proper quality assurance, and actually sit down and test out the phone with a fresh pair of eyes over the experience. Android, even on 5.1.1, is slower and more sluggish than a recent iPhone. And I say this as a devout Android user. We don’t need any more CPU cores, or any more GHz. And heck, we don’t need integrated frying pans on our phones. What we need is proper, high quality, well-tested software, written by expert embedded systems engineers. Ones who are paid to be perfectionists. Who will refuse to create a shippable version of the software until they are happy it’s just right. And that will never happen, as the commercial pressures are just too great, to get yet another phone out of the door, and move on.
The Flagship is Dying
That’s not an option for your new commodity hardware though – a replaceable battery would make your phone last too long! Much better to seal that battery up, and stop you from replacing it, to force you to buy another phone down the line. But hey, maybe by then they will have fixed that horrific memory leak, which plagues your current phone? As it’s not like you’re realistically going to get a fix for that any time soon, as the firmware developers are already having the whip cracked over them to get ready for the next generation of phones for the holidays, and for CES and MWC in early 2016.
What we need is a moratorium. A year where nobody releases a single new handset. There are already more than enough to choose from. Let’s try to get things right first! Let’s fix software on phones, and make it reliable. Let’s work out a way to separate the hardware from the OS, so we can make an operating system which is properly abstracted (like an operating system is on your PC), that doesn’t require the wheel to be reinvented each time. OEMs won’t like it – they will feel they’ll sell less phones. But if they launched less phones, and spent more time getting their existing ones to work properly, perhaps we’d actually see some real innovation on the market?
Users want a high quality experience. I can see that. You can see that. Just the OEMs can’t. They would rather churn out a new phone for the sake of it, than fix their existing ones. It’s wasteful, and it’s expensive, and it ultimately drives up their costs and prices. LG recently reported it makes an average of just 1.2 cents profit per phone they sell. That’s not a lot! It’s not sustainable.
Manufacturers are starting to wise up, and realise they can’t say the solution to every problem is to buy a new phone. That’s why it’s good news to hear that Samsung and LG will join Google in pushing monthly security updates, and working with carriers to get this to happen. Users are no longer willing to tolerate the idea of “there’s a security bug in the software on your phone, time to buy a new one.” And about time too — perhaps this will be enough to focus OEMs back onto innovating, and creating a quality experience for their customers, rather than trying to sell 2 new near-identical handsets per year, and putting off even their most loyal fans in the process, always waiting on the “next one”.
There’s always something new, just around the corner, about to hit the shelves. That’s the way of the technology industry. But given the low margins in mobile, if customers continue to feel hoodwinked by perpetual “same” product releases, and wait for the “next one”, there might not be a “next one”. Companies can no longer compete on specifications alone, and are going to instead need to innovate. A cynic could argue this explains the popularity of sealed batteries and non-upgradeable storage. But given that 2014 and 2015 flagships are incredibly capable in their own right, we could well see users entrenching themselves and waiting out for better. And heck, it’s not like you’d be suffering by doing so.
Or heck, get a glass grinder and grind the edge off the screen of your current phone, and leave it in the sun for a while. You’ve just built your own flagship! (Please don’t actually do this, or this may happen – see folks, removable batteries are important after all!)
from xda-developers http://ift.tt/1DxzXau
via IFTTT
Aucun commentaire:
Enregistrer un commentaire